PRINCETON STUDIES IN INTERNATIONAL FINANCE
No. 84, March 1998
INTERPRETING THE ERM CRISIS:
COUNTRY-SPECIFIC AND SYSTEMIC ISSUES
WILLEM H. BUITER
GIANCARLO M. CORSETTI
AND
PAOLO A. PESENTI
INTERNATIONAL FINANCE SECTION
DEPARTMENT OF ECONOMICS
PRINCETON UNIVERSITY
PRINCETON, NEW JERSEY
PRINCETON STUDIES
IN INTERNATIONAL FINANCE
PRINCETON STUDIES IN INTERNATIONAL FINANCE are
published by the International Finance Section of the
Department of Economics of Princeton University. Although the Section sponsors the Studies, the authors are
free to develop their topics as they wish. The Section
welcomes the submission of manuscripts for publication
in this and its other series. Please see the Notice to
Contributors at the back of this Study.
The authors of this Study are Willem H. Buiter, Giancarlo
M. Corsetti, and Paolo A. Pesenti. Willem Buiter is Professor of International Macroeconomics at the University of
Cambridge and a member of the Monetary Policy Committee of the Bank of England. He has previously served
at Princeton and Bristol Universities, at the London
School of Economics, and at Yale University, where he
was Juan T. Trippe Professor of International Economics.
Giancarlo Corsetti is Assistant Professor of Economics at
the University of Rome III and has also taught at Yale and
Columbia Universities. Paolo Pesenti is Assistant Professor
of Economics and International Affairs at the Woodrow
Wilson School at Princeton University. He is a research
fellow of the National Bureau of Economic Research and
has served as a consultant to the World Bank. The authors’
book Financial Markets and European Monetary Cooperation: The Lessons of the 1992–93 Exchange Rate Mechanism Crisis will be released this year by Cambridge
University Press.
PETER B. KENEN, Director
International Finance Section
PRINCETON STUDIES IN INTERNATIONAL FINANCE
No. 84, March 1998
INTERPRETING THE ERM CRISIS:
COUNTRY-SPECIFIC AND SYSTEMIC ISSUES
WILLEM H. BUITER
GIANCARLO M. CORSETTI
AND
PAOLO A. PESENTI
INTERNATIONAL FINANCE SECTION
DEPARTMENT OF ECONOMICS
PRINCETON UNIVERSITY
PRINCETON, NEW JERSEY
INTERNATIONAL FINANCE SECTION
EDITORIAL STAFF
Peter B. Kenen, Director
Margaret B. Riccardi, Editor
Lillian Spais, Editorial Aide
Lalitha H. Chandra, Subscriptions and Orders
Library of Congress Cataloging-in-Publication Data
Buiter, Willem H., 1949–.
Interpreting the ERM crisis : country-specific and systemic issues / Willem H. Buiter,
Giancarlo M. Corsetti, and Paolo A. Pesenti.
p. cm. — (Princeton studies in international finance, ISSN 0081-8070 ; no. 84
Includes bibliographical references.
ISBN 0-88165-256-3 (pbk.)
1. Foreign exchange rates—Government policy—European Union countries—Mathematical models. 2. European Union countries—Economic conditions—Mathematical
models I. Corsetti, Giancarlo M. II. Pesenti, Paolo A. III. Title. IV. Series.
HG3942.B853 1998
332.4′566′094—dc21
98-11535
CIP
Copyright © 1998 by International Finance Section, Department of Economics, Princeton
University.
All rights reserved. Except for brief quotations embodied in critical articles and reviews,
no part of this publication may be reproduced in any form or by any means, including
photocopy, without written permission from the publisher.
Printed in the United States of America by Princeton University Printing Services at
Princeton, New Jersey
International Standard Serial Number: 0081-8070
International Standard Book Number: 0-88165-256-3
Library of Congress Catalog Card Number: 98-11535
CONTENTS
1
INTRODUCTION
2
THE ERM CRISIS: A RECONSTRUCTION
Country-Specific and System-Wide Tensions in Europe
The Speculative Attacks of September 1992
Toward the Revamping of the ERM
3
BUILDING BLOCKS OF A SYSTEMIC MODEL OF CURRENCY CRISES
A Disinflation Game in a Center-Periphery Economy
The Theory of Currency Collapses—in a Nutshell
Market Expectations and Self-Fulfilling Prophecies
European Macroeconomic Spillovers: Theory and Empirical
Evidence
Currency Crises in a Multicountry Framework
1
6
6
13
25
28
28
30
33
35
38
4
THE
1992–93 EVENTS AS A COORDINATION FAILURE
Expectations Shifts in Europe During September 1992
Systemic Effects in the ERM Crisis
Cooperation at a Dead Lock?
41
42
46
49
5
CONCLUSIONS
52
APPENDIX
54
54
55
The Center Country
The Periphery Countries
Shocks to Fundamentals, Monetary Innovations, and the Real
Exchange Rate
Policy Preferences in the Periphery
Characterizing the Optimal Monetary Policy: The Shadow
Devaluation Rate
Endogenous Wages
Multicountry Nash Equilibrium
Cooperative Equilibrium
60
62
62
65
REFERENCES
66
56
59
FIGURES
1
Key Interest Rates in Germany, 1987–1995
2
Time Series of Macroeconomic Variables in Europe,
1987–1995
10
Three-Month Interest Differentials against Germany,
1990–1993
22
The Shadow Devaluation Rate
33
3
4
8
TABLE
1
Deficit and Debt Ratios in the European Union
9
BOX
1
Parameter Definitions
58
1
INTRODUCTION
The crisis of the European exchange-rate mechanism (ERM) in
1992–93 was a critical event in the post-Bretton Woods history of the
international monetary system. The disruptive effects of speculative
flows exposed the fragility of the European Monetary System (EMS)
following the removal of exchange controls, shaking the entire process
toward monetary union. Both inside and outside Europe, the events of
1992–93 represented a turning point for intellectual opinion—and
policy priorities—regarding use of the exchange rate as an effective
nominal anchor in the design of disinflation policies for integrated
capital markets.
A full understanding of the causes, origins, and implications of the
ERM breakdown can provide policy lessons that are particularly relevant,
although by no means confined, to the current debate on the monetary
future of Europe. By shedding light on the dynamics of “systemic”
crises, the lessons of 1992–93 can help to refine our interpretation of
the other two large-scale episodes of international monetary instability
occurring in the years after the European crash: the speculative attacks
and financial crises following the collapse of the Mexican peso at the
end of 1994 and the currency crises in East Asia during the second half
of 1997. Against obvious differences in their dynamics and underlying
causes, the currency crises of the 1990s are alike in sharing a rapid and
“contagious” propagation of speculative waves from the original country
or group of countries under attack to an entire region having (perceived)
comparable macroeconomic features. For this reason, an analysis of the
For their helpful comments and suggestions, we thank Michael Bordo, Barry Eichengreen, Jeffrey Frankel, Jeffry Frieden, Peter Garber, Koichi Hamada, Maurice Obstfeld,
David Pearce, Guido Rey, Kenneth Rogoff, Chris Sims, Niels Thygesen, and an anonymous referee. We also thank seminar participants at Harvard, Rutgers, Yale, Cambridge,
and Koç Universities, the University of California, Berkeley, the University of Rome III,
the London School of Economics, the Tinbergen Institute, the Banca d’Italia, the
Danmarks Nationalbank, and the European Monetary Institute, and conference participants at the Centre for Economic Policy Research conference on “Speculative Attacks on
Foreign Exchange Reserves,” held in Sesimbra, Portugal, in April 1997. Giancarlo
Corsetti gratefully acknowledges financial support from the Ministero dell’Università e
della Ricerca Scientifica e Tecnologica. We thank Mukul Kumar for excellent research
assistance. A previous version of this study has circulated as CEPR Discussion Paper No.
1466.
1
ERM crisis can enhance our tools for policy evaluation in an interdependent global economy.1
Most contributions to the recent literature have linked the ERM crisis
to the international policy conflict generated by the policy mix adopted
by Germany during its process of reunification in the early 1990s.
Explanations have also stressed the persistent asymmetric performances
of national price or unit-cost levels (reflecting divergent national
monetary and fiscal policies), the destabilizing effects of deregulating
international financial capital movements under the Single European
Act, and the perceived change in the commitment of national policymakers to fixed exchange rates after the results of the first Danish
referendum in June 1992. Other contributions have emphasized the
possible role of self-fulfilling speculative attacks triggered by sudden,
essentially arbitrary, expectations shifts in the financial markets.
Virtually all the interpretive schemes proposed by the literature have
focused on the adjustment problems faced by individual European
countries and have analyzed the country-specific sources of exchangerate tension that eventually undermined the ability of individual countries to maintain a stable parity against the deutsche mark after September 1992. In other words, most contributions have represented the
ERM as the sum of independent unilateral pegs against the currency of
the “center” country, autonomously pursued by a number of “periphery” countries.
The problem with such an approach is that it downplays or ignores
altogether the presence of structural policy spillovers in Europe, an
omission of no little consequence. Interpretations based on a countryby-country analysis cannot, by their very structure, adequately explain
the ERM events as the crisis of an exchange-rate system. They are
therefore unable to address the issue of contagious speculative attacks
or to analyze the effects of coordination (or lack thereof) of monetary
and exchange-rate policies. Theories based on country-specific imbalances (including self-fulfilling shifts in expectations) cannot easily
1
Looking backward, several contributions have also pointed to the remarkable
similarities between the ERM crisis and the international crisis of 1971–73, which led to
the demise of Bretton Woods. As Peter Kenen (1995, p. 161) writes, “in both cases, the
game played between markets and governments had ossified the exchange rate regime,
although the architects of both regimes had tried to combine short-term stability with
long-term flexibility. Furthermore, the center country in each system was unable to
arrange an orderly exchange rate realignment and resorted instead to methods that
undermined the system.”
2
rationalize the simultaneous and unanticipated eruption of financial
tensions that occurred in virtually all the countries participating in the
ERM. Models that emphasize macroeconomic developments in Germany
as the source of global shocks have little to say about why the ERM
economies, which shared comparable macroeconomic conditions,
responded to the common shock with highly dissimilar policies.
Contagious crises and systemic macroeconomic effects are central to
the reading of the 1992–93 events presented in this study. We attempt
to explain, within a comprehensive interpretive scheme, both the global
loss of credibility of the ERM parities and the subsequent asymmetric
policy developments in Europe. Espousing the view that the ERM
events derived from a “coordination failure,” we provide a model
suitable for analyzing the logic of the crisis and deriving its implications
for the behavior of key observable macroeconomic variables in Europe.
In Buiter, Corsetti, and Pesenti (1998), we generalized the theory of
currency crises so as to encompass policy links and structural spillovers
in a multicountry model of monetary coordination and financial instability. Building on those results, we analyze more directly in this study
the interplay between country-specific asymmetries and systemic issues
in the evolution of the EMS during the 1990s. The main theses discussed in our contribution may be summarized as follows.
During the first three quarters of 1992, the policy mix accompanying
the reunification of Germany, as well as sizable asymmetries in the
macroeconomic and political conditions in many European countries,
generated strong tensions, of increasing severity, in the EMS. The main
factor underlying the eruption of the ERM crisis in September 1992 was
the unwillingness by European policymakers to agree on a cooperative
policy response to these tensions. The fundamental imbalances stemming from German unification were at odds with the German-centered
model of the ERM, a model to which European economic policy had
previously conformed. Germany could not accept the idea of jeopardizing its internal price objectives, and the other countries in the system
viewed a realignment as disruptive of the policy credibility acquired
during several years of fixed exchange rates.
When, in September 1992, the existing parity grid became clearly
untenable, most European countries tried to “save” their currencies
from devaluation and refused to coordinate their exchange-rate policies. Our model predicts that, in a scenario characterized by a demand
shock in the center country of a system, a cooperative policy response
will lead to a sizable reduction in interest rates through a series of
contained devaluations of the periphery countries against the currency
3
of the center. Conversely, a noncooperative policy response—associated
with devaluations by only a subset of periphery countries—will lead to
a lower average rate of monetary expansion and to persistently higher
interest rates in the system as a whole. For this reason, the observed
devaluation rates in the periphery will be large, possibly involving
disruptive currency crises.
In 1992, the effect of the perceived policy-coordination failure on
financial-market expectations was pervasive. To the extent that the
events of September 1992 unequivocally signaled the breakdown of the
EMS as a system in which both country-specific and system-wide
monetary shocks could be counteracted by joint policy responses, the
focus of market expectations shifted away from a coordinated small
realignment to large devaluations by a few countries. Such a shift in
expectations was not the outcome of self-fulfilling prophecies but,
rather, a rational reaction by market participants to the ongoing policy
developments.
Between September 1992 and August 1993, Europe was hit by
repeated waves of speculative frenzy. Although exports from Italy and
the United Kingdom increased rapidly after the eruption of the crisis—
contributing to divert trade and employment in other countries—a
comprehensive assessment of the empirical evidence cannot overlook
two other important, systemic, features of the crisis after September
1992: the effects of the large devaluations on inflation in those countries that had maintained their peg, and, more explicitly, the financial
and real effects of the decline in the German interest rates. The
German rates had climbed relentlessly during 1991 and 1992, reaching
their highest level in the summer of 1992. After September 1992, they
fell steadily, although initially at a slow pace, through 1994. To the
extent that low inflation in the consumer-price index (CPI) and falling
interest rates allowed the nondevaluing European countries to reduce
the macroeconomic costs of their peg to the deutsche mark, the breakup
of the ERM helped, rather than hampered, the successful exchangerate policies of these countries. The novelty of our inquiry consists in
showing, within a rigorous analytical framework, how the success of
some currencies in maintaining their peg to the center currency crucially depends on devaluations by other countries, which absorb the
fundamental imbalances in the system. Although this notion runs
against the popular model of “beggar-thy-neighbor” exchange-rate policies, it stresses the particular nature of the sign of the intra-European
monetary spillovers in 1992–93 relative to the highly contractionary
monetary stance of the center country.
4
This study is organized as follows. Chapter 2 presents a historical
reconstruction of the ERM crisis. Chapter 3 describes a multicountry
model of currency crises containing several key features that suit the
stylized facts of the ERM events, that is, asymmetric macroeconomic
conditions at a country level, a global shock in the form of an aggregate
demand surge in the country at the center of the system, and structural
links involving both trade and financial dimensions. The model builds
on the traditional theory of currency collapses, developing its logical
core so as to highlight the role of international spillovers. Chapter 4
interprets the 1992–93 crisis in light of the previous chapters’ analyses.
Chapter 5 offers conclusions. A detailed, technical analysis of the
model is given in the Appendix.
5
2
THE ERM CRISIS: A RECONSTRUCTION
Country-Specific and System-Wide Tensions in Europe
From its inception in 1979, the ERM system of exchange-rate bands
allowed for periodic realignments of the central parities. During the
1980s, member countries gradually developed procedures to ensure
strict coordination of exchange-rate policies, making the setting of new
central rates a “truly collective decision” (Padoa-Schioppa, 1985,
p. 351; see also Giavazzi and Giovannini, 1989, pp. 40–41). Once a
country or group of countries proposed a realignment, the modalities
of devaluation were subject to collective negotiation in the Monetary
Committee meetings, and the final decision on implementation required
unanimous approval. The architecture of the ERM thus provided its
members with an effective institutional framework through which
economic and financial tensions arising in the exchange markets could
be absorbed through coordinated policy actions, thus anchoring market
expectations of exchange-rate realignments. At the same time, the
negotiation process prevented individual member countries from
undertaking unilateral realignments.
The eruption of the crisis in 1992, and the shifts in market expectations, coincided with clear evidence that the coordination mechanism
had somehow ground to a halt. To analyze the roots of this coordination failure, it is useful to reconstruct the ERM crisis by first focusing
on the German reunification shock and the consequent buildup of
exchange-rate tensions between the end of 1991 and the summer of
1992 (see Buiter, Corsetti, and Pesenti, 1998).
The unprecedented integration of two national economies, rebuilt
after the war under very different systems and divided by a substantial
gap in productive capacities as well as standards of living, resulted in the
adoption of a highly controversial monetary-fiscal policy mix. In 1991,
the net transfer of public funds from west to east Germany was as high
as 139 billion deutsche marks.1 A comparison to west German private
savings , which equaled 260 billion deutsche marks that year, illustrates
the magnitude of the transfer. The rate of growth of west German
1
The following year, it reached 180 billion deutsche marks. Here and throughout,
“billion” equals a thousand million.
6
public-sector indebtedness, which had averaged about 5 percent in the
second half of the 1980s, rose to 13.4 percent and 11.2 percent in 1990
and 1991, respectively (Bundesbank, 1992). The massive transfer of
resources to the east was neither financed by additional taxes nor
balanced by significant cuts in other spending items in the budget.
At the same time, with the creation of the Monetary, Economic and
Social Union in May 1990, “ost” marks could be converted into deutsche
marks on a 1-to-1 or 2-to-1 basis.2 The Bundesbank’s more cautious
proposals regarding the conversion rate of the ost mark were de facto
ignored by Chancellor Kohl. The chosen conversion rates between ost
marks and deutsche marks hit east German production hard—dropping
the level of industrial production in the first two months after reunification to less than half that of 1989—while the massive budgetary transfers
contributed to a sustained high level of consumption demand (see
Collier, 1991, and Akerlof et al., 1991).
In addition, the process of restructuring in the aftermath of reunification soon generated a noticeable increase in investment demand.
Unless much of the increase in consumer demand from the east could
be directed toward foreign goods, the German reunification scenario
would lead to domestic overheating and inflationary pressures in the
western part of the country.3 In the words of Helmut Schlesinger
(1994, pp. 6–7), “the biggest mistake in the transition which has negative consequences to this day began with the wage policy. . . . These
maladjustments were costly and they were largely responsible for . . . a
relatively restrictive line of monetary policy through the Bundesbank to
avoid creating an inflationary spiral” (for a discussion of these theses,
see Sinn and Sinn, 1992).
Publicly denouncing what its president Karl Otto Pöhl had called, in
March 1991, the “disastrous consequences” of a misguided approach to
reunification, the Bundesbank warned the German government that
reunification would provoke interest-rate increases unless the government’s budget deficit were drastically reduced. From the date of
reunification until the signing of the Maastricht Treaty in December
1991, German key rates were gradually raised, in four steps, and the
2
To be precise, all debts were converted at 2 to 1 and all claims at 2 to 1, except for
bank deposits, which were converted at 1 to 1 for a limited amount per capita (2,000
marks for children under the age of 15; 4,000 for adults under 60; 6,000 for people over
59). The average rate of conversion, according to Bundesbank estimates (1990, p. 25), was
1.8 to 1. Wage, price, and pension contracts were converted on a 1-to-1 basis.
3
Capacity utilization in west Germany reached 90 percent in 1990, and the unemployment rate fell throughout the period, dropping below 7 percent during 1990.
7
TABLE 1
DEFICIT
AND
DEBT RATIOS
IN THE
EUROPEAN UNION
(in percentages)
Deficit to GDP
Debt to GDP
1991 1992 1993 1994 1995
Austria
2.4
Belgium
6.5
Denmark
2.1
Germany
3.3
Finland
1.5
France
2.2
Greece
11.5
Ireland
2.1
Italy
10.2
Luxembourg
1.0
Netherlands
2.8
Portugal
6.5
Spain
4.9
Sweden
1.1
United Kingdom 2.6
2.0 4.1 4.4 5.5
6.6 6.6 5.3 4.3
2.9 4.5 3.9 2.1
2.9 3.3 2.5 2.3
5.8 7.9 5.5 5.0
4.0 6.1 6.0 5.0
12.3 13.2 12.5 11.4
2.2 2.3 2.2 2.5
9.5 9.6 9.0 7.8
2.5 2.1 2.3 1.4
3.8 3.2 3.0 3.3
3.3 7.1 5.7 5.4
4.2 7.5 6.6 6.2
7.5 13.4 10.4 9.2
6.1 7.9 6.5 4.2
1991
1992
1993
1994
1995
56.6 56.1 63.0 65.2 68.0
132.6 134.4 141.3 140.1 138.3
60.9 63.1 66.8 68.7 68.8
42.7 47.3 51.8 54.6 62.5
23.2 42.7 56.2 62.7 69.1
41.1 45.6 52.9 56.8 59.5
81.7 88.6 117.1 119.8 120.2
95.3 90.7 92.7 87.9 83.3
103.9 111.4 120.2 122.6 122.1
6.0
7.0
7.0
7.0
8.0
76.4 77.1 78.5 79.0 79.4
62.2 63.2 67.8 70.4 70.8
49.9 53.0 59.4 63.5 66.5
53.7 69.8 74.6 79.4 84.5
35.5 41.4 47.4 51.6 53.4
SOURCE: European Commission and OECD (for Austria, Finland, and Sweden in
1991 and 1992).
held by the Banca d’Italia suggests that speculative pressures began as
early as the spring of 1992 (Goldstein et al., 1993, p. 53). Spain and
Portugal, too, were experiencing a deterioration of competitiveness, and
in England, the strength of the pound was being undermined by a
severe contraction that generated internal pressures for a relaxation of
British monetary policy. These developments are illustrated in Figure 2,
which plots the time series of key macroeconomic variables in a number of European countries (inside and outside the ERM) between 1987
and 1995.
For the European Union (EU) as a whole, the gap between actual
public-debt and deficit performance and the Maastricht convergence
criteria widened.4 Table 1 shows how the EU members performed with
4
According to the Maastricht protocols, the general government financial deficit should
not exceed 3 percent of GDP, and its gross financial liabilities should not exceed 60
percent of annual GDP. On the transition to European economic and monetary union
(EMU) according to the Maastricht Treaty, see Fratianni and von Hagen (1992), Gros and
9
newspaper Handelsblatt issued a press release that stated: “The president of the Bundesbank, Helmut Schlesinger, does not rule out the
possibility that, even after the realignment and the cut in German
interest rates, one or two currencies could come under pressure before
the French referendum” (quoted in Muehring, 1992, p. 14). The loss of
reserves for the Bank of England was reported to be about 15 billion
dollars, almost half its entire stock. Only slightly less dramatic was the
fall of the peseta below its central ERM rate, the within-the-band
target level defended by the Banco de España.
September 16, 1992, has been nicknamed “Black Wednesday” in
England, out of respect for the wounded pride of the British monetary
authorities.7 On Wednesday morning, the Bank of England had raised
the minimum lending rate from 10 percent to 12 percent. A few hours
later, it announced a new increase to 15 percent (although it did not
implement it). At the end of the day, the pound closed in London
below its ERM floor, and in the evening, the Bank of England announced the “temporary” withdrawal of the pound from the ERM (a
few days later, on September 19, its return to the ERM was postponed
indefinitely). Later, Wednesday night, Italy followed Britain out of the
ERM, and Spain, although staying in the ERM, devalued the peseta by
5 percent. Outside the ERM, energetically and solitarily leaning against
the wind, the Riksbank pushed the marginal lending rate up to the
spectacular level of 500 percent (at an annual rate).
The following day, all the currencies that had been under attack but
had survived Black Wednesday were quoted near their ERM floors. The
Bank of England brought the minimum lending rate back down to 10
percent. The Bank of Ireland followed the Swedish recipe instead; on
September 18, it raised overnight rates to 300 percent (at an annual rate).
The French referendum on the Maastricht Treaty was finally decided
in favor of the treaty on September 20. The victory of the oui was far
from overwhelming, however (only 51.1 percent of those voting were in
favor), and speculative attacks against the franc (and also the escudo,
peseta, and punt) intensified despite the pro-Maastricht result.8 The
7
In the popular press, “black” Wednesday has recently turned “white,” reflecting the
long period of prosperity and high employment that the United Kingdom has enjoyed
since its exit from the ERM.
8
Greece, the only European Community country never to have joined the ERM, also
encountered speculative pressures on the drachma in September 1992. The Bank of
Greece raised the official lending rate from 30 to 40 percent, intervened heavily in
support of the drachma, and tightened capital controls. This last measure was reversed
toward the end of 1992.
24
two countries that had left the ERM showed no inclination to return to
the fold. On September 21, Italy announced that the lira was not
bound to rejoin the ERM in the near future, and the following day,
England cut its minimum lending rate to 9 percent.
On September 23, an unsuccessful attack was launched against the
franc. Through the interventions undertaken to contain the speculative
wave, the Banque de France suffered a loss in reserves of about 80
billion francs. The Banque de France raised the French repo rate to 13
percent, and the Bundesbank intervened heavily in support of the
franc.9 Other less evident strategies may have contributed to the
survival of the franc. It was observed, for instance, that throughout the
crisis, implicit capital controls (leading to positive offshore-onshore
interest differentials) and the implicit control of domestic lending rates
by the Banque de France (signaled by sizable differentials between money-market and prime rates) helped mitigate the repercussions of the
external financial crisis on the rest of the French economy (Marston,
1995, pp. 133–135). Formal exchange controls were explicitly introduced by Spain on September 23 and by Ireland and Portugal on
September 24.10
Toward the Revamping of the ERM
The subsequent developments of the ERM crisis can be summarized as
follows. Between October and November, tensions in the financial
markets appeared to ease. This pattern was interrupted in the second
half of November by a new financial crisis, once again originating in
the North. On November 19, the Riksbank decided to float the krona.
The impact on the weaker ERM currencies was immediate, and on
November 22, Spain and Portugal announced a devaluation by 6
percent of the central parities of the peseta and the escudo. On December 10, Norway allowed the krone to float.
Tension and relaxation alternated within the ERM during the first
quarter of 1993. Two speculative episodes are worth mentioning. On
January 4, the French franc was quoted once again near its ERM floor,
9
The speculative attack was unsuccessful, and because participants in the attack
commonly took positions for one month, most of the reserves lost were back in France by
the end of October.
10
This was not in violation of the letter of the Single Market legislation, which
retained the right for member states to impose temporary capital controls in order to deal
with disorderly exchange markets. It was, however, an additional blow to the Maastricht
timetable. Spain removed its capital controls in November 1992, after the second
devaluation of the peseta, and Portugal followed suit soon after.
25
leading the Bundesbank and the Banque de France to reiterate in a
joint statement their willingness to cooperate to defend the French
currency. The punt came under pressure in early January and was
devalued by 10 percent on January 30, despite the high interest rates
maintained throughout the month.
German interest rates were eventually cut on March 18, after the
signing of the so-called “solidarity pact,” an agreement between the
Länder (states) and the federal government about how to share the
financial burden of the reconstruction in the east. The drop in the
German official discount rate by half a percentage point to 7.5 percent
helped avert a possible new currency crisis that might have been
triggered by the neo-Gaullist sweep in the French general elections of
March 28.
In April, key interest rates fell in Germany, France, and several
other countries. These optimistic signals, however, were soon offset by
a new outburst of speculative frenzy. On April 27, the peseta was
quoted at its lowest level against the deutsche mark since Spain’s entry
into the ERM. On May 13, after a new series of heavy attacks against
the peseta, the Banco de España abandoned the defense of the ERM
band and asked for a new realignment, the third since the beginning of
the crisis. The peseta’s devaluation by 6.5 percent was followed rapidly
by an equal devaluation of the Portuguese escudo.
Good news for the ERM came from Denmark on May 18, when, in
the second national referendum, the supporters of the Maastricht
Treaty obtained a sound victory. In the weeks following the Danish ja,
the stability of the ERM was hastily (and prematurely) interpreted as a
return to the golden days of the late 1980s.
Such illusions and flights of fancy were short-lived. No later than
mid-July, the ERM was once again under pressure, following a now
familiar pattern. The French franc was quoted slightly above its ERM
floor, and French money-market rates, for the first time in six weeks,
were again above their German counterparts. Tensions had built up in
Denmark as well, forcing the Nationalbank to raise a key interest rate.
With the Bundesbank refusing to adjust its rates at its council meeting
on July 15—the last before summer vacation—the ERM reached the
point of no return.
On July 30, all the ERM currencies except the guilder and punt
were quoted at the bottom of their bands against the deutsche mark.
After an emergency weekend meeting in Brussels, a thorough revamping of the ERM was announced on August 1. Most of the surviving
“hard” ERM was replaced by a much weaker scheme for exchange-rate
26
targeting, almost indistinguishable from a free float. The size of the
bands was widened from 4.5 percent to 30 percent (15 percent on
either side of the unaltered central parities). The sole exception was
the exchange rate between the deutsche mark and guilder, the target
zone for which remained unchanged. By the end of 1993, most currencies were close to, or even above, their central rates. Yet, although
actually moving their currencies back to the previous narrow bands, the
central bankers showed no interest in legally restoring the old system.
27
3
BUILDING BLOCKS OF A SYSTEMIC MODEL
OF CURRENCY CRISES
In light of the reconstruction above, the interpretation of the 1992–93
events is by no means a straightforward task. At a minimum, there are
four fundamental questions that should not be ignored: Why did no early
warnings of the imminent turmoil emerge in the weeks preceding Black
Wednesday? Which event(s) caused the sudden, generalized shift in
expectations that triggered the first wave of speculative attacks? Why did
the crisis spread among the European economies (both inside and outside
the ERM), largely independently of the perceived strength or weakness
of their macroeconomic fundamentals? Why did the crisis last for several
months, forcing a number of countries periodically to devalue their
currencies and eventually causing a de facto demise of the system?
A model suitable for analyzing the unfolding of the EMS collapse and
of answering the above queries should be able to encompass a number of
complex features, including (1) asymmetries in the conduct of monetary
policy and in the inflation performances across countries, (2) a political
structure that attaches a credibility cost to a devaluation, and, more
crucially, (3) the presence of international externalities and structural
spillovers that tie together the policy stances in the system. The theory
described in this chapter can be seen as a step toward constructing such a
comprehensive model of system-wide currency crises.
The actual model underlying our analysis is presented in the Appendix.
We concentrate here on a verbal and intuitive discussion of the model’s
essential features, briefly relating these features to the existing literature
and emphasizing a set of results that will be especially useful in the
analysis of the 1992–93 currency collapse. 1
A Disinflation Game in a Center-Periphery Economy
Our interpretive framework considers an exchange-rate system that consists of N + 1 countries, the first N of which represents the periphery of
1
Our model builds on the familiar framework of an international monetary-policy
game. The standard references are Hamada (1976), Buiter and Marston (1985), Cooper
(1985), and Canzoneri and Henderson (1991). Recent developments are surveyed in
Currie and Levine (1993), Ghosh and Masson (1994), and Persson and Tabellini (1995).
Among recent models of currency crises, with special reference to the 1992–93 events, see
Drazen and Masson (1994), Gerlach and Smets (1994), and Ozkan and Sutherland (1994).
28
the system, while the last is the center. All countries in the system are
symmetric with respect to technology and private-sector decision rules,
including those characterizing labor-market behavior. In particular, all
countries are characterized by short-run nominal-wage rigidities: current
wages are set one period in advance, based on the expected level of the
GDP price deflator. The economies are nonetheless heterogeneous in
two dimensions: first, wage inflation rates (that is, inflation expectations)
differ across countries, and, second, policy preferences vary both within
the periphery and between the center and the periphery as a whole.
The center is characterized by its uncompromising attitude toward the
goal of domestic price stability, and its economy is not subject to systematic inflationary pressures (of the kind considered in the literature since
Kydland and Prescott, 1977, and Barro and Gordon, 1983 [1994]). In
contrast to the center, the periphery countries are characterized by
different levels of inflationary bias in policymaking. Because such distortions are present, the inflation-prone periphery countries consider fixed
exchange rates as an intermediate target on the way to price stability: the
periphery countries can choose to “tie their own hands” and reduce the
scope for discretion in monetary policy by pegging their exchange rates to
the currency of the price-stabilizing center country. Our construction
thus conforms to the “consensus” view (at the time of the crisis) of the
ERM as a disinflation mechanism, in which the center is the natural
candidate to receive the mandate of guaranteeing price stability to the
system as a whole, and the exchange rate against the deutsche mark is the
nominal anchor for the periphery countries (see Giavazzi and Pagano,
1988 [1994], Giavazzi and Giovannini, 1989, and Begg and Wyplosz, 1993).
The commitment of policymakers in the periphery is only imperfectly
credible, however, and market participants perceive the abandonment of
the peg as a possible policy option in the presence of sizable cyclical
downturns or external shocks. The concept of “imperfect credibility” of a
peg is best understood in relation to the loss of political reputation and
prestige associated with reneging on the commitment to maintain the
announced exchange-rate target: the lower the perceived political opportunity cost of devaluing or switching to a float, the lower the reliability of
the commitment to a fixed-exchange-rate regime and the credibility of
the existing parity.
Although technology and preferences underlying the economic structure
of the periphery countries are substantially identical,2 macroeconomic
2
For the sake of simplicity, the model in the Appendix assumes that all countries in the
periphery trade real goods and services exclusively with the center, that is, intraperiphery
29
stances may differ across countries: there may be wage-inflation differentials within the periphery,3 and policymakers may pursue different
objectives in terms of inflation and unemployment. Such country-specific
asymmetries are one source of exchange-rate tension in the system.
Another source of tension in the system may be a demand shock in the
center (as with German unification), generating pressures for an effective
real appreciation of the center country’s currency. In such a situation, the
periphery as a whole would benefit from a monetary expansion by the
center, which would absorb at least part of the domestic demand surge
and would lower interest rates in the system. The resulting level of
internal inflation, however, would be unacceptable to the center country,
which would benefit, instead, from a revaluation of its own currency
against the periphery, thereby offsetting the “overheating” caused by the
original domestic demand shock. Such a realignment, though, would
entail high reputational and credibility costs for the policymakers in the
periphery countries.
The Theory of Currency Collapses—in a Nutshell
To understand the implications of a realignment by a periphery country,
it is useful to follow the literature on exchange-rate crises with optimizing
policymakers (sometimes referred to as “second-generation” models of
speculative attacks) and to consider a currency crisis as a policy decision
based on the rational assessment of social and political costs and benefits.
In the Appendix, we show that such choice can be analyzed starting from
a simple optimizing condition, expressed in the metric of the exchange rate.
trade is negligible. At first sight, this last assumption may appear unrealistic. After all,
concerns about the competitive gap within the periphery have been raised frequently by
national policymakers and by both import-competing and exporting sectors in Europe
since the inception of the ERM and have affected the determination of Community-wide
exchange-rate policies. The assumption is inessential, however. A more complex setup
would not significantly alter the conclusions reached within our framework. For instance,
one of the properties of our model is that a devaluation in a given periphery country shifts
global demand toward the country’s products and causes a real appreciation in all other
periphery countries, the same qualitative result we would obtain in a model encompassing
explicitly intraperiphery competitiveness.
3
There are at least two ways in which differences in wage inflation in countries having
identical economic structures are consistent with rational-expectations equilibrium in the
model. First, if multiple instantaneous equilibria can occur (as discussed below),
expectations embodied in nominal-wage contracts may differ across countries. Second,
there may be differences in the national policy targets. Countries with a larger gap
between public and private output objectives will, other things being equal, exhibit higher
expected inflation rates.
30
This condition is based on the following two elements. First, define as
~s the “shadow devaluation rate,” that is, the percentage change in the
exchange rate when the policymaker opts for a devaluation, keeping
unchanged all other macroeconomic fundamentals. By construction, the
shadow devaluation rate coincides with the observed rate of depreciation
in a realignment; when the peg is defended, instead, the shadow devaluation rate represents the hypothetical rate of depreciation that would have
prevailed, other things being equal, if the monetary authority had decided
to realign.
Second, consider the political (or reputational) costs “paid” by the
policymaker when reneging on its commitment to defend the existing
exchange-rate parity. These costs are better understood as a proxy for the
wide array of nonquantifiable political interests underlying the defense of
a given exchange-rate target, ranging from naked national chauvinism to
fears of loss of political prestige, reputation, and influence. They may also
reflect a widespread belief that exchange-rate stability is a public good in
its own right, quite apart from its anti-inflationary implications. 4 We
denote these utility costs by c.
Indexing the (logarithm of the) current exchange rate and the current
fixed parity with s and s–, respectively, it will generally be possible to find
some increasing function, g(.), such that the equilibrium exchange-rate
behavior can be characterized as follows:
(1)
This condition expresses the optimal choice of a policymaker who carries
out a social-welfare arbitrage across monetary-policy regimes: the policymaker will defend the peg as long as the gains from monetary discretion—in
terms of higher employment, output, and consumption—are low compared
to the reputational or political cost of a devaluation (that is, ~
s < g[c]); it
will optimally abandon the peg otherwise.
For a deeper understanding of the implications of the policy rule
(equation 1), consider the links between our analysis and the so-called
“first-generation” models of speculative attacks subsequent to Salant and
Henderson (1978) and Krugman (1979 [1992]). 5 The concept of a shadow
4
In more complex models based on repeated games between private and public agents,
these costs would represent the endogenous reputational penalty associated with the policymaker’s deviation from the optimal outcome; see, for example, de Kock and Grilli (1993).
5
See the survey by Garber and Svensson (1995). Buiter, Corsetti, and Pesenti (1995,
1998) and Cavallari and Corsetti (1996) generalize the analytical framework to the new
generation of models of currency crises allowing for optimizing policymakers.
31
exchange rate was introduced into the literature by Flood and Garber
(1984 [1994]). In the context of these seminal models, monetary policies
are exogenous (that is, not explicitly derived in terms of welfare optimization), and a devaluation involves no reputational cost, so that c = g(c) = 0.6
The mechanism leading to a currency crisis is the maximization of speculative profits by forward-looking private agents, who understand the
fundamental inconsistency between a policy of fixed exchange rates and a
trend in domestic credit. In such a framework, the shadow devaluation
rate is a measure of private profits per unit of foreign reserves.
In our framework, that is, with optimal endogenous policies, the shadow
devaluation rate becomes a measure of the opportunity cost, in socialwelfare terms, of defending the existing parity. A devaluation occurs when
the net welfare benefits associated with the abandonment of the peg, s~ ,
outstrip the reputational costs, g(c). Thus, the shadow devaluation rate is
both a measure of the welfare gains from switching regimes and the
optimal rate of devaluation that the policymaker actually implements
when abandoning the peg. The use of the shadow devaluation rate helps
simplify the exposition of the analysis, while offering important insights
about the basic economic mechanism at work.
In the context of our model (see Appendix), the shadow devaluation
rate of each of the N countries in the periphery is a simple linear function
of the (logarithms of the) wage rate, the price and employment objectives
of the government, and a country-specific shock (to be discussed below).
Denote such a shock by Li , where the subscript i refers to a particular
country in the periphery. Also, denote by w the (logarithm of the) wage
level in domestic currency, with o being an index of domestic policy goals
(for example, price and employment targets), and p* being the price level
of the center country. Then, the shadow depreciation rate in the periphery country i is simply
s~ = )w + *o + +L p*
(2)
i
i
i
i
where ), *, and +, are positive constants depending on the parameters
of the model. In models with optimizing policymakers, a devaluation is the
optimal reaction of the government to the fundamentals faced by the
country (indexed by L i , o i , and p* in our model) and to market expectations
(indexed by the forward-looking wage rate, wi , in our model).
A graphical synthesis of our analysis appears in Figure 4. On the x axis,
we put the support of the shock Li. Conditional on given values of wi , oi ,
6
In principle, a positive c could capture the transactions costs of carrying out financial
arbitrage.
32
are also the determinants of the “credibility” of the peg. Focusing, for
instance, on the domestic variables that appear in equation (2), high wageinflation rates or overambitious policy goals (that is, a large level of the
index o) increase the likelihood of a crisis by raising the shadow devaluation rate for each level of the shock.
Forward-looking wages, w (and, in general, all macroeconomic
variables embodying market expectations), play a complex role in the
second-generation models of currency collapses. Market prices are set
based on an assessment of the probability of a devaluation; the choice
whether to devalue or not, however, depends, ceteris paribus, on observed price and wage rates. This circular structure of the expectation
game between the private and public sectors may make devaluation
expectations self-fulfilling, that is, validated ex post by the actual behavior of the monetary authorities.
It follows that, when the political costs of a devaluation are not
excessively high, there might be different forecasts of future monetary
policies, all consistent with rational expectations. Market participants
may coordinate on a “good” equilibrium, expecting stable exchange
rates and setting a low level of wage inflation, or on a “bad” equilibrium,
expecting a realignment and setting a high level of wage inflation. Many
studies have extensively explored the predictions of this class of models,
introducing additional elements so as to account for reputation-building
devices by the policymaker, as well as for the interplay between fiscal
and exchange-rate policies.7
In terms of our model, consider two economies facing the same countryspecific fundamentals, that is, the same shock and similar policy
targets o. These two economies, which are structurally identical, may
display quite different macroeconomic outlooks. One may have low
wages, low prices, and high competitiveness; because its shadow devaluation rate is below the devaluation threshold g(c), its currency is strong.
The other may have relatively high wages and price inflation, shifting
the shadow devaluation rate to the right for any level of the fundamentals and lowering the credibility of the peg.8
7
See, for example, the contributions by Obstfeld (1986, 1991, 1994, 1995), Jeanne
(1995), and Velasco (1996), and the discussions in Kenen (1996) and Buiter, Corsetti, and
Pesenti (1998). Eichengreen, Rose, and Wyplosz (1994) provide an empirical analysis of
speculative episodes, with emphasis on the ERM crisis, contrasting the predictions of firstand second-generation models of currency crises. Bordo and Schwartz (1996) survey
historical case studies of currency crises in relation to the two alternative interpretive
approaches.
8
Models based on self-fulfilling prophecies have so far been unable to provide a
convincing theory of the coordination of individual expectations on a particular equilibrium.
34
European Macroeconomic Spillovers: Theory and Empirical Evidence
Relative to the number of analyses that build on the considerations of the
previous section, few contributions on the ERM crisis have satisfactorily
analyzed the international determinants of domestic credibility. A point
commonly made is that the stability of the peg in the periphery is enhanced by inflation in the center—in our specification, a high p* shifts
the shadow devaluation rate to the left in Figure 4. The conflict between
the center and the periphery as a group, however, is not the only element
of policy interaction across countries: a second dimension of strategic
links is found in intraperiphery spillovers. Because these spillovers play a
crucial role in our discussion, this section will analyze their nature and
implications in some detail.
Consider a monetary expansion in one periphery country (for instance,
Italy), which brings about a real depreciation of its currency against the
center (Germany).9 The impact of such a monetary expansion on the
economy of another (any other) periphery country (say, Belgium) can be
split into two components of opposite sign, an expenditure-switching
effect and an expenditure-changing effect.
These effects are best understood starting from the equilibrium condition in the goods market of the center country (Germany), holding
constant this country’s level of production. The real depreciation of the
periphery country’s currency (the lira) will appreciate the terms of trade
of the center country, and, because output supply in the center is kept
constant, equilibrium will require a fall in the center’s real interest rate.
If all else is equal, the real appreciation of the center’s currency will shift
demand in the center away from the goods produced in either the center
or the rest of the periphery (that is, Belgium) and toward Italian goods.
This is the expenditure-switching effect associated with the Italian devaluation, which lowers aggregate demand in the rest of the periphery. The
fall in the center’s real interest rate, however, will lower the real interest
rate in all other periphery countries and will boost demand for Belgian
goods. This is the expenditure-changing impact of a monetary expansion
in Italy, which increases global demand by lowering the system-wide
interest rate.
The relative likelihood of a specific equilibrium is generally modeled by specifying an
exogenously given, and essentially arbitrary, distribution function defined over all possible
outcomes.
9
With predetermined wages, a country’s GDP deflator increases with its nominal
money stock, but less than proportionally, and its real output expands. The increase in
aggregate demand that matches the increase in supply requires a real depreciation and a
fall in the real interest rate.
35
If, following a monetary expansion by any one periphery country, the
elasticity of the aggregate demand is large enough with respect to the real
exchange rate—that is, if the expenditure-changing effect prevails over
the expenditure-switching effect, the real exchange rate will appreciate
and the real interest rate will fall in all other countries in the periphery.
Arguably, such a scenario describes well the economic links among
European countries in the period of the crisis. Nonetheless, it is logically
possible that the spillovers associated with a monetary expansion by Italy
will have the opposite sign: all else being equal, a nominal devaluation of
the lira may shift global demand away from the goods produced in the
rest of the periphery, resulting in a real depreciation of all periphery
countries, along with Italy, as well as in a generalized rise of the real
interest rates in these countries.
Despite the implausibility of such a scenario in 1992, whether expenditure-switching effects prevail over expenditure-changing effects or vice
versa is ultimately the kind of question that can only be addressed by
using a multicountry econometric model. The problems in carrying out
empirical estimates of these effects are well known, and there is considerable disagreement in the empirical literature about the sign and size of
international spillovers (see, for instance, Frankel, 1988, and Ghosh and
Masson, 1994, chap. 2). In the discussion that follows, we review the
findings of several contributions on structural links in Europe.
Both the International Monetary Fund’s MULTIMOD model (Masson,
Symansky, and Meredith, 1990) and the MSG model (McKibbin and
Sachs, 1991)—based on a Mundell-Fleming framework incorporating
rational expectations, asset dynamics, and intertemporal budget constraints—provide strong support for our hypothesis that a German monetary expansion increases output and decreases real interest rates in the
periphery. The available empirical evidence on the intraperiphery spillovers is less compelling, because few studies in the existing literature
provide an evaluation of intra-EU policy multipliers.
Masson, Symansky, and Meredith (1990) simulate the effects of a 10
percent increase in the U.K. money-supply target on France, Italy, and
Germany, starting in 1990.10 They find that the U.K. monetary expansion
has a positive effect on French and Italian outputs both on impact and
after a one-year lag and that the impact on German output is non10
The authors consider Britain as the only EMS, non-ERM, country, and they do not
include France, Italy, and smaller industrial countries in their assessment of monetary
multipliers, because “monetary policies in these countries are constrained by the need to
limit deviations of their exchange rate vis-à-vis the Deutsche mark” (Masson, Symansky,
and Meredith, 1990, p. 23).
36
negative.11 The effects on short-term interest rates are negligible in the
periphery, although German interest rates fall by ten basis points in the
two years following the U.K. monetary expansion.
McKibbin and Sachs (1991) consider the effects on the EMS countries
of a monetary expansion in the “rest of the OECD” (the OECD countries
minus Japan, the United States, and the EMS countries). Assuming that
the policy multipliers for the “rest of the OECD” represent an acceptable proxy for the policy multipliers for the subset of ERM countries that
adopt an expansionary stance, real interest rates at any maturity are
found to be persistently lower, relative to the baseline, in both Germany
and the “rest of the EMS” following the increase in the money supply.
The effects on output are temporarily negative but become positive after
a one-year lag.
More recently, Hughes Hallett and Ma (1995) use an updated version
of MULTIMOD to run a number of simulation exercises on costs and
benefits from coordination for the period from 1990 to 1996. Particularly
interesting from our vantage point is the simulation of a “no realignment”
scenario in which the United Kingdom and Italy target, respectively,
DM 2.95 ± 2.5 percent per pound and DM 1.33 ± 2.5 percent per 1,000
lira after 1992. The results are unambiguous. The monetary tightening
associated with the defense of the Italian and U.K. currency bands leads
to a generalized increase in short-term real interest rates in both Germany and France: “Interest rates have to rise to levels not seen since the
late 1970s. Real interest rates are very high therefore. That generates a
persistent recession” (Hughes Hallett and Ma, 1995, p. 32; see also
tables 3 and 4, pp. 46–47).
We believe that the results of the simulation exercises based on multicountry econometric models can be approached only with necessary and
appropriate reservations—perhaps with some degree of skepticism. It is
worth noticing, however, that the simulation results discussed above
confirm the widely held view that the ERM devaluations in 1992 substantially contributed to the fall of interest rates both in Germany and in the
rest of the system (France and the United Kingdom, in particular) between 1992 and 1993 and do not support the presence of substantial
negative spillover effects. This conclusion has been reinforced by the 1995
report of the Commission of the European Communities (1995, paras.
9–11, 23) on the impact of currency fluctuations on the Internal Market:
11
The percentage deviations of real GNP from baseline are, for France, 0.4 in 1990 and
0.2 in 1991; for Italy, 0.3 and 0.2; and, for Germany, 0.2 and 0.0 (Masson, Symansky, and
Meredith, 1990, table 16, p. 33).
37
Changes in nominal exchange rates have had differing effects on cost competitiveness, depending on the country in question. In some cases, they have offset
the cost-competitiveness gains or losses recorded over the period from 1987 to
1992. In others, they have led to net gains or loss in cost competitiveness
compared with the trends over the period. . . . In conclusion, it should be
emphasized that there is no strict relationship between variations in nominal
exchange rates and variations in cost competitiveness. . . . It transpires that
exchange rate fluctuations have not had a significant impact on current account
balances. . . . Trade movements measured over the entire period 1987–94 show
that the exchange rate fluctuations recorded since 1992 have not hitherto been
of major importance at the aggregate level.
In summary, although the size and sign of macroeconomic spillovers in
Europe are difficult to measure and are, perhaps, not stable over time,
the pattern of prevailing expenditure-changing effects fits well the empirical evidence for the period under consideration. To the extent that the
impact of the “swing” of German interest rates on the countries of the
European periphery is considered the predominant macroeconomic issue
in the unfolding of the ERM events, the net effect in the transmission of
monetary policies in Europe can be deemed, on balance, to be positive.
This assumption does not rule out the possibility of intraperiphery shifts
in demand from strong- to weak-currency countries’ goods following a
devaluation; it only deemphasizes their relevance for domestic welfare
compared to the impact of a system-wide fall in real interest rates.
Currency Crises in a Multicountry Framework
The presence of externalities and policy spillovers posits a clear challenge
to analyses of the 1992–93 crisis that implicitly treat the ERM as the sum
of independent and inward-looking unilateral pegs against the deutsche
mark. The key insight of our approach is that the modalities of a currency
crisis in any given country, as well as the characteristics of the new
macroeconomic equilibrium on which the country settles after the crisis,
cannot be properly assessed without accounting for macroeconomic
developments and policy options in the rest of the system. Black Wednesday sheds light on the 1992 attack on the French franc, and the behavior
of the franc helps explain the crisis of the peseta.
As seen above, in our discussion of currency collapses, it is the realization of the country-specific shock, i , that ultimately drives the
decisions of a national policymaker trying to minimize his (or her)
country’s own loss function. In the multicountry model presented in the
Appendix, the shock i includes two components (others allowing for
38
domestic-productivity shocks and velocity shocks could easily be added).
A first disturbance term, indexed by ,, depends on current and future
anticipated shocks to the aggregate demand of the center country (that is,
the IS shock). This affects all periphery countries symmetrically. A second
component of the country-specific shock depends, with a negative sign, on
the monetary innovations of all the other countries in the system (including the center). By indexing the monetary stances with m* (for the center
country) and mj (for country j in the periphery), we can write:12
Li / , N(m w) >2j (mj wj )
j…i
N > 0 , >2 > 0 .
(3)
In equilibrium, an increase in demand for the center’s output requires,
ceteris paribus, a real effective appreciation of the center’s currency
against the periphery. This can be achieved by (a) domestic inflation in
the center, (b) an appreciation of the nominal exchange rate of the center
against the periphery, or (c) a generalized deflation in the periphery.
Under option (a), the demand boom in the center would be absorbed
internally and would have no significant consequence in the periphery:
heuristically, the changes in , and m* would offset each other in equation
(3), leaving Li unaffected everywhere in the periphery. Moreover, as considered before, a monetary expansion in the center would also increase p*,
decreasing the shadow devaluation rate (equation 2) for any level of L i.
If the center is unwilling to tolerate the inflationary consequences of
the internal demand boom, the shock will be transmitted to the periphery.
If the periphery consists of one country only, for any given domestic
monetary policy (option b), the IS shock in the center will unambiguously
lead to CPI inflation in the periphery. The mechanism is well known. As
the demand boom raises the interest rates in the center, capital will flow
out of the periphery country. Holding the money supply constant, the
equilibrium exchange rate of this country will depreciate in both nominal
and real terms against the center, and its CPI level will increase. If,
instead, the monetary authorities of the periphery react to the fall in the
demand for domestic currency and defend the current parity, the corresponding fall in domestic money supply will have a recessionary impact
(option c).
In the general case of a multicountry system, equation (3) shows that,
from the vantage point of each individual periphery country, the impact of
12
Note that the country-specific shock depends on money supply in excess of the
nominal wage rate. To the extent that wages are set based on monetary expectations, that is,
that wt = Et-1 mt , only monetary surprises matter in determining Lt .
39
the shock originating in the center depends not only on the policy response
of the center itself, but also on the monetary behavior of the rest of the
system. In other words, although the relevant shock i is always country
specific, it reflects the behavior of all the other countries in the system.
With positive spillovers ( > 0), the role played by a monetary expansion
in the rest of the periphery is somewhat analogous to the role played by a
monetary expansion in the center. By bringing the center’s effective terms
of trade closer to their equilibrium level, an expansion somewhere in the
periphery represents a “shock absorber” for the rest of the system.13
So, although the domestic shadow devaluation rate completely characterizes the monetary policy of a country, it is not possible to determine an
individual country’s shadow devaluation rate without knowing the policy
stance of the other countries in the system. This implies that the institutional characteristics that affect the average monetary stance in the
exchange-rate system are a key determinant of the individual currency
crises. In the next chapter, we shall turn to the characterization of the
system-wide equilibrium and show that, even if the relation between the
center and periphery is frozen in noncooperative behavior, the macroeconomic outcome is still dependent on whether and to what extent currency
crises are mitigated through cooperative policy actions among the periphery countries themselves.
It should be emphasized that our approach does not rule out the
possibility of multiple equilibria for given country-specific fundamentals,
but we do not base our interpretation of the expectations shifts that
trigger a currency crisis on self-fulfilling prophecies. We provide, instead,
a coherent framework within which to interpret the ERM demise in terms
of fundamentals only, provided that the traditional list of fundamentals is
augmented by the rules of the international monetary-policy game, that is,
the cooperative or noncooperative design of monetary and exchange-rate
policies in the system.
13
When we move from the partial-equilibrium approach considered so far to generalequilibrium considerations, it should be clear that the policy stance of every country in the
system is jointly determined as a function of the exogenous shock to fundamentals .
40
4
THE 1992–93 EVENTS AS A COORDINATION FAILURE
The idea that a policy-coordination failure was at the root of the ERM crisis
underlies most interpretations that stress the role of the German unification shock in the 1992–93 events. According to this view, the origins of
the crisis must be traced to the international policy conflict between the
center country, which was unwilling to bear the inflationary consequences
of its internal unification, and the periphery countries, which were unwilling either to sustain the costs of a deflation or to accept the political
consequences of a devaluation (see Eichengreen and Wyplosz, 1993). As
Tommaso Padoa-Schioppa (1994) emphasizes, this conflict prevented
member countries from agreeing on a common objective and a common
policy in the course of 1992 and ultimately compromised the normal
operating procedures in the ERM system at the time of the crisis:
[The cause of the ERM crisis] was plainly traceable to what in the academic
jargon is called a “co-ordination failure.” . . . There was the refusal to accept a
general realignment and even to call a meeting of the Monetary Committee or
of the ministers and central-bank governors when, in September 1992, a
general realignment might have calmed the markets. The general procedure,
once embarked on, did not produce a credible new grid. At various times, and
in various ways, through unhelpful declarations that excited markets as well as
through policy decisions that caused unnecessary friction, the system was
destabilized by its very custodians (Padoa-Schioppa, 1994, pp. 14–15).
Although rhetorically suggestive, the notion of a coordination failure
among the “very custodians” runs the risk of appearing vague and, perhaps, tautological; almost by definition, a crisis of an exchange-rate
system is a symptom of insufficient or ineffective policy coordination.
To interpret the ERM crisis in the context of a coordination failure, it is
necessary to provide a consistent analysis of the implications of a cooperation breakdown for the modalities of a currency crisis. Only within
the framework of a multicountry model of currency collapses (such as
the one developed above) is it possible to define rigorously the concept
of “coordination shock” and to explore in depth the predicted responses
of financial and real markets to such an event.
The analysis that follows in this chapter provides the theoretical
underpinnings of two theses. First, the eruption of the 1992 crisis was
associated with a discrete change in market beliefs about intra-ERM
41
policy cooperation, prompted by the modalities of the first ERM realignment on September 14. Second, the instability of the European exchangerate system and the contagious speculative waves in the aftermath of
Black Wednesday were to a large extent the outcome of fundamental
systemic imbalances, as is predicted by a model of currency crises among
noncooperative, interdependent economies. The chapter concludes by
discussing the hypothesis that, relative to the difficulties of coordinating
monetary policies on anything but a strenuous defense of the current
parities, the coordination failure in September 1992 could be interpreted
as a rational choice by ERM member countries.
Expectations Shifts in Europe During September 1992
Weathering the crisis through monetary coordination. Recalling our
reconstruction in Chapter 2, we note that two important “stylized facts”
characterize the period leading up to the 1992–93 crisis. First, the proposal
for a coordinated realignment formulated by Germany and Italy during the
weekend preceding the crisis included a modest nominal revaluation of
the deutsche mark (3.5 percent), to be coupled with an equally modest
devaluation of the lira (3.5 percent). Whether or not such a realignment
could have “saved” the ERM is an open issue. For our purposes, it is
sufficient to point out that the proposal, supposedly designed to serve as
a basis for further negotiation among the ERM countries, did not include
drastic changes of the existing parities.
Second, although the credibility of the ERM was diminishing in August
1992, all indicators of devaluation expectations—such as interest-rate
differentials, forward premiums, and average forecasts based on survey
data—were by no means large by historical standards (see Rose and
Svensson, 1994). This piece of evidence is typically interpreted in one of
two ways, either that irrationality and myopia were affecting market
forecasts or that, in September 1992, there was a sudden (and substantially
arbitrary) shift in market sentiments consistent with the existence of
multiple rational-expectations equilibria.
Our theoretical construction allows us to put forth a different and comprehensive interpretation of the factual evidence. As late as the summer of 1992,
both European policymakers and financial markets were still anticipating
some cooperative defense of the integrity of the exchange-rate system.
Expectations of a coordinated, generalized realignment led market participants to forecast a relatively small adjustment of the ERM parities.
Facing an exogenous demand shock in the center, , consider the
implications of a coordinated realignment that includes a revaluation of the
center country’s currency against all other currencies in the system. An
42
example would be the German-Italian proposal after the Bath meeting, a
scheme that, to our knowledge, is the only plan ever brought forward as
the basis for the design and implementation of a coordinated realignment.
What would be the system-wide macroeconomic outlook after such a
revaluation? The answer to this question is rooted in the international
spillovers of domestic monetary policies.
From our previous discussion, we know that intra-European externalities
are positive, in the sense that a monetary expansion by one country reduces
the impact of global shocks and raises social welfare in the rest of the system.
A devaluation by one country decreases CPI inflation abroad and contributes
to the loosening of the global monetary stance. This decreases real interest
rates and therefore boosts aggregate demand in the system as a whole.
With positive externalities, cooperation calls for “doing more” in terms
of monetary policy. In other words, the overall monetary stance is more
expansionary in an equilibrium with a coordinated realignment than in a
noncooperative equilibrium. The very fact that many countries join the
realignment scheme implies that there will be a substantial effective
devaluation of the periphery against the center.
The main point here is that, from the vantage point of each individual
periphery country, the relatively large expansionary stance in a coordinated
realignment need not require a large devaluation of the periphery country’s
own currency. Because it is the average revaluation of the center’s currency that ultimately matters, individual rates of nominal devaluation in the
periphery need not be high for purposes of domestic stabilization when
interest rates fall everywhere in the system (see Appendix and Buiter,
Corsetti, and Pesenti, 1998, chap. 8).
It is worth emphasizing that these results do not require that the center
coordinate its monetary policy with the periphery. The presence of
externalities and policy spillovers implies that, even though the center is
unwilling to act cooperatively, the periphery countries may still increase
their total welfare by coordinating monetary policies among themselves.
Of course, if one allows for center-periphery coordination contingent on
the periphery countries accepting a realignment, then the average devaluation by the periphery that is necessary to reestablish equilibrium in the
system will be even smaller. In equilibrium, an additional monetary
expansion by the center (above the monetary expansion consistent with its
noncooperative behavior) reduces devaluation rates for all currencies (see
our discussion at the end of Chapter 3).
The implications for our reconstruction are straightforward. To the
extent that market participants in the late summer of 1992 based their
realignment expectations on the presumption that the ERM would
43
remain an effective coordination device, they rationally forecast that a
new parity grid involving a large number of small realignments would be
the most likely policy response to global and country-specific tensions in
Europe. The anticipated interest-rate cut in Germany, in parallel with the
realignment, reinforced these expectations.
Expectations under a noncooperative scenario. The next step toward
understanding the effects of a coordination failure on market expectations
is the characterization of an alternative equilibrium—different from the
cooperative scenario described above—under the assumption that the
proposal for a revaluation of the center is rejected and that each country
unilaterally maximizes its own objective function taking monetary policy
pursued in the rest of the system as given. Referring the reader to the
Appendix for details, we describe below two essential features of such a
noncooperative equilibrium.1
First, when the periphery countries do not coordinate their monetary
and exchange-rate policies against the center, the equilibrium response to
shocks to the center demand, , may require that only a subset of countries devalue their currencies. If some countries devalue and therefore
help to lower the real interest rates in the system, the other countries will
have less incentive to abandon the peg. In a noncooperative equilibrium,
some countries may avoid realignment altogether.
All things being equal, the number of countries that optimally choose
to devalue will increase in response to both the demand shock to the
center and to the cumulative misalignment in the system (as indexed by
the average magnitude of w s–). If the combined effect of these shocks
is very large, all countries will devalue in equilibrium. Thus, by focusing
on uncoordinated realignments including a subset of countries as equilibrium outcomes, we implicitly rule out scenarios in which the size of the
exogenous shocks is excessively large.
Second, because countries ignore the positive externalities associated
with their money expansion, the average monetary stance is always less
expansionary than in the cooperative scenario. The realignment has
therefore only a limited impact on the equilibrium real interest rate. But
this result implies that the individual devaluation rates of the countries
that abandon the peg may be significantly larger in a noncooperative
equilibrium. Facing persistently high system-wide interest rates, countries
1
In the Appendix, we first characterize the optimal policy of a country that abandons
the peg, as opposed to a country that maintains the peg. We then assume a particular size
and composition of the set of countries that abandon the peg. Finally, we check whether
the conjectured set satisfies a system-wide equilibrium condition.
44
that realign may find it optimal to engineer sizable devaluations; that is,
they may be subject to currency crises.
In light of these considerations, market expectations based on the
noncooperative scenario are remarkably dissimilar across countries. Each
country is expected to determine whether it is preferable to peg or to
devalue, and to determine this independently on the basis of the costbenefit analysis described in the previous chapter (that is, by comparing
the country-specific shadow devaluation rate with the devaluation threshold). Excluding the extreme cases in which the shock to center demand is
either too large or too small, a number of periphery countries will be
expected to abandon the peg and to devalue dramatically in response to
fundamental disturbances while the other periphery countries will be
expected to maintain the defense of their exchange-rate parities. In
equilibrium, periphery countries facing the same global shock will be
expected to pursue asymmetric policies somewhat independently of their
domestic macroeconomic positions.
The lira devaluation shock. In September 1992, a sudden shift in
devaluation expectations in a number of ERM countries triggered the first
wave of speculative attacks. The question is which unexpected event led
market participants to revise their beliefs. We suggest that it was the
realignment of the Italian lira on September 14, 1992.2
The fact that only the lira was devalued that day—against expectations
of a realignment to restore equilibrium in an interdependent monetary
system and amidst rampant rumors and revelations about fundamental
disagreements among European policymakers—gave the financial markets a strong signal about the ongoing policy game in Europe. With no
substantial concession by the Bundesbank, a new parity grid including a
mere 7 percent depreciation by only one currency could not appear
credible as the outcome of cooperative policy.3 It should be recalled that
the lira devaluation against the deutsche mark on September 14 was
identical in size to that proposed two days earlier in the context of a
general realignment, which would have had all other countries devaluing
their currencies by 3.5 percent against the deutsche mark.
2
In support of this interpretation, see BIS (1993, pp. 198–199), quoted in Kenen (1995,
pp. 159–160).
3
As Kenen (1995, p. 159) observes, “the 7 percent devaluation of the lira on September
14, 1992, was badly bungled. It may have been large enough to restore Italy’s competitive
position, but it was too small to seem decisive. It evoked an inadequate response by the
Bundesbank—a 25 basis-point cut in the discount and Lombard rates. And it should have
been the occasion for a general realignment—a modest revaluation of the deutsche mark
combined with selective devaluations of the lira, peseta, escudo, and pound.”
45
In describing the noncooperative scenario of our model, we have
pointed out that, if the exogenous shocks to fundamentals are not too
large, restoring equilibrium will require sizable devaluations by a subset
of countries in the system. To the extent that the modalities of the lira’s
realignment modified market views about the level of cooperation in the
system, expectations of exchange-rate devaluation were correspondingly
revised. In other words, after the first Italian devaluation, it became
apparent that German rates would fall significantly only in response to
sharp devaluations by a number of other countries. Ultimately, an analysis
of the crisis as a policy-coordination failure can rationalize the dramatic
change in the financial markets’ assessment of ERM stability in the week
of the lira realignment without resorting to “sunspots” and the exogenous
uncertainty of the self-fulfilling-prophecies theory.
Systemic Effects in the ERM Crisis
In the new, “uncoordinated” scenario that emerged after the lira shock,
the first wave of speculative attacks hit a set of currencies that appeared
weak with regard to conventional indicators. These currencies included—
in addition to the lira—the escudo, peseta, pound, and punt. Some observers were puzzled by the attack against the French franc on September
23 and asserted that the traditional arguments emphasizing the role of
weak fundamentals in triggering a currency crisis did not apply to the
French case. The puzzlement grew when the franc came again (and
repeatedly) under pressure during the 1992–93 period. These “paradoxical” attacks on the franc have been proposed as a key piece of empirical
evidence supporting the view that self-fulfilling prophecies were at the
root of the ERM crisis.
The nature of the attacks against the franc is a highly controversial
subject in the literature. Taking issue with the “self-fulfilling” view, some
observers have pointed out that the empirical evidence at the time of the
crisis did not entirely support the contention that the French fundamentals were strong; in particular, the high level of unemployment in 1992
was a clear indicator of the French economic malaise. In the light of our
model, we can go beyond the terms of this debate. Even if the French
macroeconomic conditions were indeed relatively sound, the overall
strength of the French fundamentals could not be assessed independently
of the monetary behavior of the other countries in the system.
Disregarding strategic interactions (that is, using a partial-equilibrium
perspective), one may be tempted to predict that, in the presence of
currency-market tensions generated by a shock in the center, countries
with ex ante weaker macroeconomic conditions will be the most likely
46
candidates for a currency devaluation, and the equilibrium devaluation
rate will be “proportional” to the magnitude of domestic imbalances. A
general-equilibrium analysis, however, shows that such a prediction may
be proved wrong. According to the logic of a systemic model, a “strong”
domestic performance—as measured by the standard set of macroeconomic indicators—is no guarantee against the possibility of a currency
crisis.
Because of international interdependencies, neither the number of
countries that devalue nor which countries devalue is uniquely determined in a noncooperative scenario. For a given exogenous shock to
center demand, , equilibria that include more currencies in a realignment tend to be characterized by lower individual devaluation rates.
Certain equilibrium outcomes, moreover, cannot be ruled out: countries
with relatively high inflation—or out-of-line policy targets such as misaligned exchange-rate parities—may be able to keep their pegs, provided
that a sufficiently large group of periphery countries participate in a
(noncoordinated) realignment. By the same token, there may exist
equilibria in which countries with relatively high cost competitiveness
(as measured by unit-labor cost at the current parity) devalue their
currencies.4 Ultimately, policy spillovers weaken the relation between
domestic macroeconomic conditions and exchange-rate stability.
Restoring international equilibrium after a sizable shock to the center
requires a sizable fall in the system-wide real interest rate and a sufficiently
large real appreciation of the center’s effective exchange rate against the
periphery. Both real interest rates and exchange rates are functions of the
average size of the realignment in the periphery, that is, of the average
monetary stance in the system. Absent cooperation, if not enough countries
have already devalued by a sufficient amount to restore system-wide
equilibrium, both “weak” currencies (such as the escudo, the peseta, and
the pound) and “strong” currencies (as the French franc was perceived to
be) may be candidates for a speculative attack.
An important implication of this analysis is that the magnitude of
devaluation rates in the ERM was to some extent the mirror of the
success of a few countries (notably France) in defending their currencies.
The conventional wisdom attributes negative spillover effects (beggar-thy4
Consider equations (2) and (3) for a country with weak domestic macroeconomic
conditions (wi and oi). Heuristically, the effects of devaluations by other countries on the
country-specific shock i may more than compensate for weak domestic levels of wi and oi,
lowering the shadow devaluation rate below the threshold g(c). Provided enough countries
devalue in equilibrium, country I will be able to keep the peg, despite its weak macroeconomic conditions.
47
neighbor) to a devaluation. According to the logic of the traditional approach, the ERM realignment raised the cost of defending the currencies
that did not participate in it and therefore increased the likelihood of
speculative attacks against them. This logic downplays the effects of a
devaluation by one country on the inflation rates and interest rates of
countries that do not devalue.
In our interpretation, the noncoordinated devaluation of a subset of
currencies partly reduced the ERM-wide imbalance and therefore helped
to defend those currencies that had not been devalued. In the Europe of
1992–93, burdened by sizable real interest rates, the contagion effect
from a currency crisis was therefore opposite to the traditional prediction.
It was not the prospect of losing employment and trade to the devaluing
countries that increased the likelihood of a currency crisis in ERM
countries that had not devalued. Rather, it was the adverse effects on
interest rates caused by the strenuous defense of the existing parities by
the nondevaluing countries.
This proposition clearly implies that the size of devaluation rates in
Europe responded to more than domestic imbalances such as inflationrate differentials, budget deficits, and unemployment rates. On empirical
grounds, the prediction of our model is consistent with the results of
econometric analyses of the ERM, which, despite the wide variety of
model specifications proposed, find no link between these “domestic
fundamentals” and the likelihood and magnitude of currency devaluations
(see Eichengreen, Rose, and Wyplosz, 1994).
It is worth pointing out that most empirical work on this issue—both
econometric studies and noneconometric research agendas such as the
analysis of early-warning indicators of currency collapses—typically
builds on the logical structure of unilateral-peg or two-country models of
currency crises. The findings of our multicountry model offer a simple
interpretation of the empirical evidence resulting from a unilateral-peg
approach: systemic effects make the link between domestic macroeconomic conditions and the size of currency devaluations, let alone the
likelihood of a crisis, tenuous. Because internal and external macroeconomic developments are strictly interwoven, this link may even take the
“wrong” sign.
Our claim finds support in recent econometric work focused on the
contagious effect of currency crises both inside and outside the EMS. The
findings by Eichengreen, Rose, and Wyplosz (1996), for instance, show
that, after controlling for a number of fundamental determinants of the
exchange rate, the probability of a speculative attack against one currency
is significantly affected by currency crises and exchange-rate tensions
48
elsewhere in the world. This econometric evidence is consistent with the
prediction of our “fundamentals” model of systemic exchange-rate crises
and reinforces the view that unilateral-peg models focusing exclusively on
domestic macroeconomic variables may provide a misleading framework
for policy analysis.
Cooperation at a Dead Lock?
At the root of our interpretation of the 1992–93 events is the thesis that
European countries were unable to cooperate in defense of the ERM.
Understanding why such a failure in coordination occurred may be a more
difficult task than analyzing its positive implications. Economic theory
suggests a set of stylized reasons why rational agents may fail to coordinate their actions. These include a lack of commitment technologies,
disagreement about the distribution of costs and benefits from common
actions, and a lack of appropriate policy tools by which to redistribute
gains across individuals. The breakdown of ERM cooperation in 1992 can
be interpreted in terms of a combination of these three elements, with
particular emphasis on the difficulties of reaching an agreement about the
distribution of costs and benefits from a policy that might have counteracted the tensions in the system.
The German authorities in 1992 were unwilling to lower interest rates
and to reduce the financial pressure on the ERM countries independently of a realignment. But their position “disregarded” the presence
of country-specific political costs of a devaluation in the periphery,
which hindered the possibility of reaching consensus on the modalities
of devaluation. It is worth stressing, in this respect, that, according to
the standard procedures regulating the collective decisions on
exchange-rate policies in the ERM, a realignment required the unanimous approval of all member countries.
Should national policymakers have devalued simultaneously (that is,
should the periphery have accepted a revaluation of the center) and
tolerated the consequences of a cumulative loss of prestige and credibility?
If not, which countries should have paid the reputational cost of a devaluation in the general interest? Should individual devaluation rates have been
asymmetric, so as to account for the ex ante heterogeneity across countries (different rates of wage inflation)? If so, how large should individual devaluations have been? Should the countries that maintained the
peg have been compensated by the devaluing countries for suffering a
deterioration of cost competitiveness? Or, rather, should the devaluing
countries have been compensated for the loss of anti-inflationary credibility? The (formidably complex) answers to these questions would have
49
determined how the costs and benefits from periphery-wide policies
were to be distributed among the individual periphery countries.
Note that, even if periphery countries could have agreed on a specific
distribution of costs and benefits from a selective and (or) asymmetric
realignment, they still lacked the tools required to achieve such a distribution. In the interdependent EMS of 1992, these tools should have taken the
form of improbable redistributive and compensatory transfers intended to
correct the perceived unfairness of particular joint-policy measures.
These considerations suggest that in the presence of a center that pursued a realignment as a way to absorb its internal imbalance (somewhat
irrespective of the credibility costs in the rest of the system), the set of
politically feasible coordination schemes was very small, so small, perhaps,
that the only feasible coordination in September 1992 was on the status
quo, that is, on a painfully deflationary defense of the existing parities.
Political constraints on coordination thus made it impossible to pursue
a cooperative monetary relaxation in the system to compensate for the
German policy mix.. In such a scenario, it was perhaps both individually
and collectively rational to abandon attempts to find a joint response to
the existing imbalances. Uncoordinated behavior, through large devaluations by a few periphery countries, was the only way to loosen up the
average monetary policy in the system.
For most ERM countries, once the crisis had started, the survival of the
parities was clearly linked to individual efforts, with little or no intraCommunity support. The difficulty of the task was magnified by a widespread feeling that the strength of the political support for Maastricht had
dwindled with the first Danish referendum. From September 1992 until
August 1993, the operation of the ERM was characterized by uncoordinated attempts to determine the new equilibrium exchange rates, that is,
by the (somewhat messy and staggered implementation of the) Nash
scenario of our model.
The two obvious exceptions to this pattern are the Netherlands and
France. The Netherlands was substantially a member of the European
center—its currency never came under attack and its monetary policies
were fully harmonized with Germany. France, although clearly not a
member of the “hard” center for the period under discussion, held a
special position in the extended center or “soft core” of the system.
Several observers have commented on the evidence of a coalition or
cooperative arrangement between Germany and France, and unlike the
other periphery countries, which were effectively left to fend for themselves, France benefited from massive German support in the defense of
its parity.
50
As mentioned above, the particular position of France had important
implications for the dynamics of adjustment in 1993. Although the repeated waves of speculative attacks on the franc can be explained in terms
of systemic imbalances in Europe (a devaluation of the franc would have
contributed substantially to the appreciation of the effective deutsche
mark rate), the successful defense of the French franc implied that the
long-term interest rates in Europe remained relatively high, causing other
European currencies to experience repeated realignments or sustained
depreciations. The analysis applies to ERM currencies such as the escudo,
peseta, and punt, as well as to non-ERM currencies. In the period under
consideration, high interest rates were creating difficulties for the Scandinavian countries that kept their currencies pegged to the deutsche mark,
thereby exacerbating already severe country-specific problems (for
Finland, the breakup of its trading arrangements with the former Soviet
Union; for Sweden, the collapse of domestic consumption demand).
According to our interpretation, the final act in the ERM crisis—the
widening of the fluctuation bands in August 1993—represented explicit
recognition that the EMS had lost its ability to be a coordination device
for the domestic policies of its members. The width of the bands accurately reflected the weakness of international ties in what remained of
the monetary system in Europe, and the end of the “hard ERM” represented the beginning of a new stage in European monetary relations.
The policy coordination of the past, with its strict and problematic
requirements in terms of design, consultation, and implementation, was
replaced by the goal coordination of the present, that is, the attempt to
pursue identical policy targets (the convergence criteria) according to
the letter of the Maastricht or Amsterdam Treaty. From this vantage
point, any form of monetary cohabitation that will emerge in Europe is
bound to reflect—for good or for evil—the implications of the coordination failures of the early 1990s and to represent, ultimately, the persistent inheritance of the ERM crisis.
51
CONCLUSIONS
Building on a model of intra-European spillovers consistent with the
available empirical evidence, this study has developed a coherent interpretation of the two main aspects of the ERM crisis: the sudden change in the
regime of expectations in September 1992 and the domino effects of
speculative attacks between the fall of 1992 and the summer of 1993. Our
model of an integrated EU economy, which allows for both countryspecific macroeconomic conditions and global shock, has focused on the
systemic forces linking the revision of market expectations and asymmetric
policy responses as different dimensions of the same equilibrium outcome.
On methodological grounds, our theoretical framework has analyzed the
implications of changes in the degree and modalities of policy coordination
on the equilibrium exchange-rate parities in Europe. We have shown how
the traditional theory of currency crises can be generalized in a multicountry framework in which the fundamentals are augmented so as to
include the rules regulating coordination (or competition) among policymakers. We have explained how, in equilibrium, the interdependence
among policymakers weakens the link between domestic macroeconomic
conditions and the strength of a currency. The effect of a relatively strong
macroeconomic outlook on the stability of a currency may be offset by
policy shocks from abroad, reflecting the ongoing policy game among
national central-bank and fiscal authorities. By the same token, relatively
weak currencies may not be forced to devalue when global shocks disturb
the system, if a sufficient fall in the world interest rate absorbs the initial
imbalance.
The revamping of the ERM in 1993 has closed an important chapter of
European monetary history. The belief, popular throughout the 1980s,
that the exchange rate is an effective nominal anchor and that countries
can import credibility for their disinflation strategies by pegging their
currencies to a price-stabilizing center country, is nowadays met with
strong reservations and widespread skepticism (see Svensson, 1994, and
Obstfeld and Rogoff, 1995). In particular, since 1993, the academic and
policy debate has focused mainly on the idea that a flexible peg—that is, a
fixed-exchange-rate regime with escape clauses—may be destabilizing and
may contain the seeds of its own disruption.
Although we certainly sympathize with this view of unilateral pegs, we
believe that an additional, important lesson can be learned from the
52
1992–93 ERM crisis. Macroeconomic disturbances and shifts in market
sentiments need not be disruptive of international exchange-rate arrangements. The speculative tensions in Europe since 1992 have been to a large
extent the reflection of each country’s attempt to restore macroeconomic
equilibrium within a new EMS suffering from the loss of its systemic
structure and nature. The widening of the ERM bands in 1993 represented the de facto institutional sanctioning of such a transformation.
Since 1993, the wide-band ERM has been a largely empty shell, in which
European countries have pursued noncooperative policies while trying to
rebuild a common structure.
In light of this conclusion, the future evolution of EMU depends
critically on the ability of the new European institutions to provide
effective solutions to the issues that prevented the system from surviving
in 1992, that is, their ability to reestablish the conditions for an effective
cooperative framework. The ERM collapsed when European policymakers could not agree on the distribution among the member countries
of the costs and benefits from policy interventions. Similarly, the case can
be made that, regardless of the speed and sustainability of intra-European
macroeconomic convergence, monetary cohabitation in Europe will be
severely jeopardized by the widespread perception that EMU imposes
unfair obligations on, and grants asymmetric benefits to, its members. In
this sense, a crisis would revive the as yet unreconciled political faults that
were exposed in the system by the 1992–93 crisis.
53
APPENDIX
This appendix presents the structure of the model underlying our analysis.
All variables other than interest rates are in natural logarithms. Variables
referring to the center country are starred. Variables referring to the periphery countries are indexed with a subscript i, for i = 1, 2, ..., N . Unless
otherwise stated, Greek letters (both lowercase and uppercase) refer to constant, positive parameters.
The Center Country
Output supply in the center, denoted by y∗ , is a function of employment,
n∗ , subject to decreasing returns to scale:
yt∗ = (1 − α)n∗t
0<α<1.
(A.1)
Profit-maximizing competitive firms equate the marginal product of labor
to the real wage. The money wage in the center is denoted by w∗ , and the
center’s GDP deflator is denoted by p∗ :1
wt∗ − p∗t = −αn∗t .
(A.2)
Real aggregate demand in the center depends on the effective real exchange rate of the periphery relative to the center z (defined below), the
center’s real interest rate r∗ (defined below), and an aggregate demand shock
λ∗ :
yt∗ = λ∗t − δzt − νrt∗ .
(A.3)
The center’s real exchange rate is defined as follows. Let si be the
nominal spot exchange rate of the periphery country i (expressed as country
i’s currency per unit of center’s currency), and let pi be the GDP deflator
of the periphery country i (in local currency). The real exchange rate of
the periphery country i relative to the center is defined as
zi,t = si,t − pi,t + p∗t .
(A.4)
Under the assumption of symmetry, the effective real exchange rate of the
periphery relative to the center z is then simply given by the arithmetic
average of the real exchange rates in the periphery:2
zt ≡
N
1 zi,t .
N i=1
(A.5)
1 Strictly speaking, defining the logarithm of the nominal wage as w
∗ , the demand for
∗ − p∗ = ln(1 − α) − αn∗ . For notational simplicity, we define it in
labor is given by w
∗ − ln(1 − α).
the following discussion as w∗ ≡ w
2 A simple, albeit algebra-intensive, modification of the model can account for differences in country size or trade shares by modeling the effective exchange rate as a
weighted average of the bilateral rates.
54
Assuming a constant share of imports in consumption, β (which applies
to each of the periphery countries as well as to the center), the center’s CPI
is defined as follows:3
qt∗ ≡ (1 − β)p∗t + β
N
1 (pi,t − si,t ) = p∗t − βzt
N i=1
0<β<
1
.
2
(A.6)
The real interest rate in the center is its nominal interest rate, i∗ , minus
the expected proportional rate of change in its CPI, q ∗ :
∗
rt∗ ≡ i∗t − Et qt+1
+ qt∗ ,
(A.7)
where Et denotes the expectation operator conditional on information available in period t.
Assuming a constant-velocity money-demand function, equilibrium in
the money market requires
m∗t = p∗t + yt∗ = wt∗ + n∗t ,
(A.8)
where m∗ denotes the center’s nominal money stock. At the end of period
t − 1, that is before the center money stock m∗t is determined and observed,
wage setters choose the money wage prevailing in period t. Their objective
function is to minimize the forecast deviation of employment from the fullemployment level (here normalized to zero). Therefore, they solve
min
Et−1 (n∗t )2 ,
∗
wt
(A.9)
subject to (??). Because n∗ = m∗ − w∗ , this implies that nominal wages
are equal to the expected money supply, and employment (or output) is a
function only of monetary innovations:
wt∗ = Et−1 m∗t
(A.10)
n∗t = m∗t − Et−1 m∗t .
(A.11)
The Periphery Countries
Periphery countries have the same technology as the center. Thus, using
self-explanatory notation, the supply-side equations characterizing the periphery are given as
yi,t = (1 − α)ni,t
(A.12)
3 We restrict the propensity to import β to be less than one half. As will become
clear later, this assumption rules out the possibility that real interest differentials and
real expected depreciation between center and periphery move in opposite directions.
55
wi,t − pi,t = −αni,t .
(A.13)
We assume that periphery countries import (export) goods and services
exclusively from (to) the center country. This is the reason why only the
bilateral real exchange rate of country i relative to the center, zi , enters
into the demand equation for country i’s output:
yi,t = λ + δzi,t − νri,t .
(A.14)
In contrast to its role in the demand equation in the center country (??),
the parameter λ is constant in equation (??). In other words, for the sake
of simplicity, we abstract from country-specific and time-specific IS shocks
hitting the periphery countries. The other behavioral parameters δ, ν, and
β are identical in both the center and the periphery.
Real interest and exchange rates in country i are
ri,t = ii,t − Et qi,t+1 + qi,t
(A.15)
qi,t = pi,t + βzi,t .
(A.16)
By analogy with the center, real-money balances, money wages, and employment in the periphery are determined as follows:
mi,t − pi,t = yi,t
(A.17)
wi,t = Et−1 mi,t
(A.18)
ni,t = mi,t − Et−1 mi,t .
(A.19)
Finally, we assume that assets denominated in different currencies are
perfect substitutes in private agents’ portfolios, so that the uncoveredinterest-parity condition holds:
ii,t = i∗t + Et si,t+1 − si,t .
(A.20)
Note that, given (??), with perfect capital mobility, the uncovered-interestparity condition must hold for any pair of currencies in the system.
Shocks to Fundamentals, Monetary Innovations, and the Real Exchange Rate
In this section, we present a semi-reduced form of our model, expressing all
endogenous variables as functions exclusively of exogenous, predetermined,
or control variables. Consider, first, the bilateral real-interest-rate differential between country i and the center country, ri,t − rt∗ . By taking the
sum over the N periphery countries, the average interest-rate differential
between the periphery and the center will be
i ri,t
= rt∗ + (1 − 2β)(Et zt+1 − zt ) .
(A.21)
N
56
According to the previous expression, the real-interest-rate differential and
the expected rate of depreciation of the real exchange rate between center
and periphery move in the same direction if and only if β < 1/2, that is, if
there is home bias in consumption preferences (??).
Using (??) together with the aggregate demand functions (??) and (??)
and the resource constraint of the economy as a whole, defined as
∗
∗
i yi,t
i ni,t
(A.22)
− yt = (1 − α)
− nt ,
N
N
we obtain a first-order stochastic-difference equation in zt :
φ
i ni,t
− n∗t +
(λ∗ − λ) ,
zt = γEt zt+1 + φ
N
1−α t
(A.23)
for which the parameters γ and φ are defined in Box 1.
Because the effective real exchange rate z is a forward-looking variable,
we impose a no-bubble terminal condition. Solving (??) with such a boundary condition yields
∗
i ni,t
(A.24)
− nt + t ,
zt = φ
N
where is defined as
∞
φ s
γ Et (λ∗t+s − λ) .
t ≡
1 − α s=0
(A.25)
The effective real exchange rate depends both on the difference between
the current monetary innovations in the periphery and in the center and,
through the variable , on the present discounted value of current and
expected future real-demand shocks in the center relative to the periphery.
Thus, a demand (IS) shock in the center larger than in the periphery causes
the center’s real exchange rate to appreciate, whereas a money-supply shock
in the center larger than in the periphery causes the center’s real exchange
rate to depreciate.
Next, it is straightforward (though algebraically tedious) to show that
the bilateral real exchange rate of periphery country i relative to the center
is
zi,t = ξ (1 − θ) ni,t + υi,t ,
(A.26)
where θ and ξ are defined in Box 1. Note that the sign of θ is ambiguous.
In (??), as well as in (1) in the main text, υ denotes the relevant macroeconomic disturbance from the vantage point of country i, which is a function
57
BOX 1
P D
v(1 − 2β)
<1
2δ + v(1 − 2β)
γ
≡
φ
≡
θ
≡
ξ
≡
φ
1−α
=
>0
δ + ν(1 − β)
1 − Nθ
ρ
≡
α + βξ(1 − θ)
ρc
≡
α + βφ
κ
≡
α + ξ(1 − θ)
κc
≡
α+φ
Σ
≡
1 + σρ2
κ
Σc
≡
1 + σ (ρc )
κc
Λ
≡
σρ
Λc
≡
σρc
A
≡
(Σ − Λ) /Σ
Ac
≡
(Σc − Λc ) /Σc
B
≡
1/Σ
Bc
≡
1/Σc
C
≡
(Σ − βΛ) /Σ
Cc
≡
(Σc − βΛc ) /Σc
Γ
≡
ξθ
C<1
ξθ + κ
1−α
2δ + v(1 − 2β)
1
δ − νβ
2δ + ν(1 − 2β) N
58
2
of the global shock and the policy stances of all other countries in the
system:
nj,t = t
υi,t ≡ t − φn∗t − ξθ
j=i

− φ (m∗t − wt∗ ) + ξθ

(mj,t − wj,t ) .
(A.27)
j=i
Adopting the notational simplifications of Box 1, we can finally write
the semi-reduced-form equations for the CPI as
qi,t = ρni,t + wi,t + βυi,t ,
(A.28)
and the bilateral nominal exchange rate relative to the center as
si,t = κni,t + wi,t − wt∗ − αn∗t + υi,t .
(A.29)
Policy Preferences in the Periphery
The policymakers in the periphery country i minimize an intertemporal loss
function defined as follows:
∞
Et Li,t+τ ,
(A.30)
τ =0
where the single-period loss function is
Li,t ≡ i,t + ci Ii,t ≡
1
(ni,t − n̄i )2 + σ(qi,t − q̄i,t )2 + ci Ii,t
2
0 if si,t = s̄i,t
.
Ii,t =
1 otherwise
(A.31)
(A.32)
The single-period loss function i in (??) is quadratic in the deviation
of actual employment and CPI from their current target levels, n̄ and q̄,
respectively. The target levels for prices and employment, as well as the
exchange-rate parity (indexed by s̄) are known at time t − 1, before wages
are set.4
The target level of employment exceeds the rational-expectations equilibrium of “natural” level (normalized to zero in this model): n̄ > 0. Following the standard conventions, such a parametrization of the model implies
4 Over time, the CPI target level q̄ and the nominal-exchange-rate target s̄ may change,
but not independently of each other. See the discussion in Buiter, Corsetti, and Pesenti,
1998, chaps. 6-7.
59
the presence of exogenous (and unremovable) distortions in the periphery
labor market, which make the full-employment output level socially suboptimal. The well-known theoretical implication of the resulting conflict
between public preferences and equilibrium constraints is that an equilibrium with full monetary discretion is affected by an inflationary bias.
The positive constant c in (??) denotes the welfare cost of abandoning the peg: country i’s policymakers suffer a welfare loss equal to c when
the current exchange rate deviates (no matter by how much) from the
announced exchange-rate parity. Such cost indexes the “commitment technology” of the monetary authority: the higher the degree of commitment to
the defense of the peg, the higher the cost, c, that the policy authority will
pay (in terms of reputation and credibility) if it abandons its announced
target.
Characterizing the Optimal Monetary Policy: The Shadow Devaluation Rate
In the absence of international policy coordination, the optimal policy rule
for a periphery country combines two different monetary regimes. In one,
the money stock is consistent with the survival of the peg. In the other, the
peg is abandoned and the money supply optimally responds to fundamentals. If the policymaker decides to defend the current parity s̄, the money
supply is endogenous and is implicitly determined by (??), where we posit
s = s̄. For future reference, define the employment level, the CPI, and the
loss function implied by s = s̄ as nF X , qF X , and F X , where FX denotes
“conditional on defending the peg.”
If, instead, the current exchange-rate parity is no longer a binding target
or constraint, the policymaker will choose a monetary policy that minimizes
the current loss function, . By taking the first-order condition as a minimum, it can be easily shown that the optimal devaluation rate (contingent
on a realignment) is
∆s̃i,t = Awi,t + (1 − A) q̄i,t + Bn̄i + Cυi,t − s̄i,t − p∗t ,
(A.33)
where the parameters A, B, and C are defined in Box 1. Posing oi ≡
n̄i + (1 − A) /Bq̄i − s̄i /B, we obtain the expression (2) in the main text.
Equation (??) defines the shadow devaluation rate, that is, the percentage difference between the (optimally chosen) value of the exchange rate
and the target exchange rate if the peg were abandoned. Note that the
shadow devaluation rate is increasing in the predetermined nominal wage,
in both the employment and the price targets, as well as in the countryspecific shock.
By definition, the prevailing exchange rate conditional on the abandonment of the peg will be s = s̄ + ∆s̃. We define the employment level, the
CPI, and the loss function implied by s = s̄ + ∆s̃ as nF L , qF L , and F L ,
60
where FL denotes “conditional on abandoning the peg.” Note that the
shadow devaluation rate can also be written as
κ FL
L
FX
FX
q − qi,t
=
.
∆s̃i,t = κ nF
i,t − ni,t
ρ i,t
(A.34)
The interpretation of these relationships brings additional insights to the
meaning of ∆s̃. The shadow devaluation rate is proportional to the “employment gap” (nF L − nF X ), that is, the employment loss caused by defending the existing parity. It is also proportional to the “price level gap”
(q F L − qF X ), that is, the inflation benefit gained by defending the peg. In
either case, the shadow devaluation rate provides a measure of the welfare
opportunity cost of maintaining the exchange rate as fixed.
To emphasize the relevance of the latter point, consider the policymaker’s choice between the two exchange-rate (and monetary-policy) regimes.
The policymaker will opt for abandoning the peg if and only if the loss
under a peg is larger than the loss associated with a devaluation (including
the lump-sum welfare cost c):
si,t = s̄i,t
X
FL
if F
i,t − i,t ≤ ci
si,t = s̄i,t + ∆s̃i,t
X
FL
if F
i,t − i,t ≥ ci .
(A.35)
By using (??) and the first-order condition for minimizing , we can rewrite
the condition for an optimal choice of exchange-rate regime (??) exclusively
in terms of the shadow devaluation rate:
si,t = s̄i,t
if ∆s̃i,t ≤ ci
si,t = s̄i,t + ∆s̃i,t
if ∆s̃i,t ≥ ci ,
(A.36)
where c̃ is a constant obtained as a transformation of the original devaluation cost c, that is, c̃ = g(c) for some monotonic function g.5 In other
words, there exists a threshold value of the shadow devaluation rate that
triggers an optimal realignment.6 It bears emphasizing that such a threshold c̃ translates the opportunity cost of abandoning the peg from the metric
of the welfare function into the metric of the exchange rate.
5 In our case, the constant can be shown to be equal to the square root of
2cκ2 / 1 + σρ2 .
6 Note that the escape clause specified in our analysis does not preclude a priori the
possibility of a revaluation of the central parity. We simplify the analysis by ruling out
construction shocks to fundamentals that would correspond to a large negative value of
the shadow depreciation rate (so that ∆s̃ ≥ −c̃). The extension to the general case is
simply a corollary of the analysis that follows.
61
Endogenous Wages
To determine the equilibrium wage rate, we proceed as follows. First, because the shadow devaluation rate is monotonic in υ, there exists a threshold
level for the shock, say ῡ, such that ∆s̃ ≥ c if and only if υ ≥ ῡ. We shall
refer to the event υ ≥ ῡ by the shorthand FL, and to the event υ ≤ ῡ by
the shorthand FX.
Second, define the probability of a realignment as
πi,t ≡ Pr{∆s̃i,t ≥ c̃i } ≡ Pr {υi,t ≥ ῡi,t } .
(A.37)
Wage-setters’ forecasts will be obtained by taking the expectations of mF X
and mF L , conditional on, respectively, the defense of the peg (FX ) and a
devaluation (FL), and combining them according to their respective probabilities:
X
wi,t = Et−1 mi,t = (1 − πi,t ) Et−1 mF
i,t |F X
L
+ πi,t Et−1 mF
(A.38)
i,t |F L .
The wage rate obtained according to this expression will be a decreasing
function of the devaluation threshold ῡ, so far taken as an exogenous parameter.
Third, we now replace w in the definition of the shadow devaluation rate
∆s̃ with expression (??). Rearranging the realignment rule (??), it can be
shown that country i will devalue its currency if the following condition
holds:
Bn̄i + (1 − A) (q̄i,t − s̄i,t )
1 − Aπi,t
A [(1 − πi,t ) Et−1 (υi,t |F X) + πi,t (1 − C) Et−1 (υi,t |F L)]
−
.
1 − Aπi,t
c̃i ≤ Cυi,t − p∗t +
(A.39)
This expression is critical to the endogenous identification of the devaluation threshold ῡ, so far taken as a given parameter. The equilibrium
interior value(s) of the devaluation threshold under rational expectations
can be found by taking expression (??) to hold with equality and solving
for υ = ῡ.
Multicountry Nash Equilibrium
For the sake of simplicity, we shall assume, in what follows, that the devaluation costs c are equal across countries. In addition, the center implements
a noncontingent money rule,
m∗t = 0 ,
62
(A.40)
so as to rule out any strategic interaction between center and periphery as
a whole. Although permitting us to avoid a great many analytical complications, such a simplification does not affect the results of our analysis in
any substantive way.7 It can be easily verified that the center’s monetary
policy implies
n∗t = m∗t − Et−1 m∗t = 0 ,
(A.41)
as well as p∗t = wt∗ = 0.
To simplify the notation, we drop the time subscripts. In order to
determine the Nash equilibrium, we proceed in three steps. First, for any
country j, the realized rate of exchange-rate depreciation can be written as
∆sj ≡ sj − s̄j = (ξθ + κ) nj + (wj − s̄j ) + η ,
where
η = − ξθ
N
nj .
(A.42)
(A.43)
j=1
This second equation is obtained by rearranging (??) as a function of η,
that is, the global shock net of the system-wide monetary response (including country j response). Note that the level of η will be endogenously
determined in equilibrium.
Second, consider the shadow
devaluation rate of country i as a function
of j=i nj . To solve for j=i nj , sum (??) across all j = i and rearrange.
After substituting into (??), we obtain
∆
si = Hi + K − Γ
∆sj ,
(A.44)
j=i
where Γ is defined in Box 1. The shadow devaluation rate is a function of
three elements. The first element is an index of country-specific asymmetries (including potential asymmetric wage rates arising from self-fulfilling
changes in expectations):
Hi = [Boi + Awi + Γ (s̄i − wi )] .
(A.45)
The second element is an index of net periphery-wide shocks, labor-cost
imbalances, and policy spillovers:
K = C + Γ
(wj − s̄i ) + Γ(N − 1)η .
(A.46)
j
7 Indeed, we obtain comparable results in our analyses of international monetary
games allowing for a quadratic loss function in employment and the CPI level for both
the center country and the periphery countries; see Buiter, Corsetti, and Pesenti (1995,
1998).
63
The third element is the aggregate devaluation rate by all other countries in
the periphery of the system. Under our assumptions about the sign of the
external effects of domestic monetary policy, this term is negatively related
to country i’s shadow devaluation rate.
Next, consider a realization of the shock that is not “too large” and
assume the existence of an equilibrium in which F number of countries
devalue and N − F countries maintain the peg, with F to be determined
endogenously. The set of devaluing countries will be indexed by FL, which
stands for “floater.” After some simple manipulations of the above formulas, we can write two expressions for the equilibrium shadow devaluation
rate, one for a representative floater in the set FL in (indexed i):
(1 − Γ)K − Γ i∈FL Hi
1
∆
si = ∆si =
Hi +
1−Γ
1 + Γ(F − 1)
≥ c
∀i ∈ FL ,
(A.47)
the other for a representative “pegger” (indexed z):
(1 − Γ)K − Γ i∈FL Hi
∆
sz = Hz +
≤
c
∀z ∈
/ FL .
1 + Γ(F − 1)
(A.48)
For each country in the set FL, the first expression yields the equilibrium devaluation rate as a function of the number of countries that devalue
in equilibrium and the index K, which depends on the global monetary
stance of the periphery. Once η is endogenously determined, for a given
level of the global shock , the two expressions jointly determine the number of countries that devalue in equilibrium, F . This number must satisfy
the following inequality for all Hi of the floaters and all Hz of the peggers:
C + Γ j (wj − s̄j ) − ΓF
Hi
c − Hz
A H̄F +
c−
≤κ
≤
,
(A.49)
1−Γ
F Γκ + (α + φ)(1 − Γ)
1−Γ
where H̄F is the average of Hi for all floaters i ∈ FL.
Note that the number F of countries abandoning the peg in equilibrium increases with the size of the global shock and with the cumulative
labor-cost imbalance w − s̄ (an index of exchange-rate misalignment) in
the periphery. In equilibrium, each country in the set FL finds it optimal
to abandon the peg, and each country outside the set finds it optimal to
maintain the announced exchange-rate parity, provided that the countries
in FL optimally devalue. If countries are ex ante identical (Hi = H̄F ), it
will not be possible to determine which particular countries will be part of
the set FL. If the countries are not too heterogeneous, the same result will
obtain: if the average devaluation rate in the system is sufficiently high,
64
countries with relatively weak fundamentals may not devalue in equilibrium, and countries with relatively strong fundamentals may be part of the
set FL.
Note also that, for given , Hi , and Hz , there can be more than one
value of F satisfying the above inequality. In this case, the smaller the
value of F that solves (??), the larger will be the corresponding individual
devaluation rates.
The range of shocks for which some countries keep the peg in equilibrium is limited by the fact that F must lie between 0 and N , the total
number of periphery countries. The boundaries of this range can thus be
determined by setting F = 0 and F = N in equation (??) and solving for
the corresponding threshold values of , say min and max . These thresholds correspond to the largest Hi for the floaters, and the smallest Hz for
the peggers. For shocks larger than max , all countries in the periphery
will devalue by their specific optimal rate ∆s̃ > c̃. For shocks smaller than
min , all periphery countries will maintain the peg.
Cooperative Equilibrium
In the cooperative equilibrium described in the main text, the policymakers
in the periphery minimize the joint-loss function,
∞
N
1 Et Li,t+τ ,
(A.50)
N τ =0
i=1
with respect to the N monetary stances. The first-order condition with
respect to the monetary instrument of country i is
ni − n̄i + σρ (qi − q̄i ) = σβξθ
(qi − q̄i ) .
(A.51)
j=i
Aggregating, the average devaluation rate (as a result of the coordinated
realignment) is
j ∆sj
j wj
j q̄j
j n̄j
j s̄j
c
c
c
c
=A
+ (1 − A )
+B
+C −
, (A.52)
N
N
N
N
N
where the parameters Ac , Bc , and C c are defined in Box 1. Comparing this
equation with its analog in the absence of cooperation (??), it is straightforward to show that C c > C. The average policy response to a global
shock is more expansionary under a joint cooperative realignment than in
the absence of cooperation, even if all countries find it optimal to devalue
in a noncooperative equilibrium.
65
REFERENCES
Akerlof, George A., Andrew K. Rose, Janet L. Yellen, and Helga Hessenius, “East
Germany in from the Cold: The Economic Aftermath of Currency Union,”
Brookings Papers on Economic Activity, 1 (1991), pp. 1–87.
Bank for International Settlements (BIS), 63rd Annual Report, Basle, Bank for
International Settlements, 1993.
Barro, Robert J., and David Gordon, “Rules, Discretion, and Reputation in a
Model of Monetary Policy,” in Torsten Persson and Guido Tabellini, eds.,
Monetary and Fiscal Policy, Vol. 1: Credibility, Cambridge, Mass., and London, MIT Press, 1994, pp. 99–121 (previously published 1983).
Begg, David, and Charles Wyplosz, “The European Monetary System: Recent
Intellectual History,” in Centre for Economic Policy Research, The Monetary
Future of Europe (report of the conference held on December 11–12, 1992, in
La Coruña, Spain), London, Centre for Economic Policy Research, 1993.
Bini-Smaghi, Lorenzo, Tommaso Padoa-Schioppa, and Francesco Papadia, The
Transition to EMU in the Maastricht Treaty, Essays in International Finance No.
194, Princeton, N.J., Princeton University, International Finance Section,
November 1994.
Bordo, Michael D., and Anna J. Schwartz, “Why Clashes between Internal and
External Stability Goals End in Currency Crises, 1797–1994,” National Bureau
of Economic Research Working Paper No. 5710, Cambridge, Mass., National
Bureau of Economic Research, August 1996.
Buiter, Willem H., Giancarlo M. Corsetti, and Paolo A. Pesenti, “A CenterPeriphery Model of Monetary Coordination and Exchange Rate Crises,”
National Bureau of Economic Research Working Paper No. 5140, Cambridge,
Mass., National Bureau of Economic Research, June 1995.
———, Financial Markets and European Monetary Cooperation: The Lessons of
the 1992–93 ERM Crisis, Cambridge, Cambridge University Press, 1998.
Buiter, Willem H., Giancarlo M. Corsetti, and Nouriel Roubini, “Excessive
Deficits: Sense and Nonsense in the Treaty of Maastricht,” Economic Policy, 16
(April 1993), pp. 57–100.
Buiter, Willem H., and Richard C. Marston, eds., International Economic
Policy Coordination, Cambridge and New York, Cambridge University Press,
1985.
Canzoneri, Matthew B., and Dale W. Henderson, Monetary Policy in Interdependent Economies: A Game-Theoretic Approach, Cambridge, Mass., MIT
Press, 1991.
Cavallari, Lilia, and Giancarlo M. Corsetti, “Policy Making and Speculative
Attacks in Models of Exchange Rate Crises: A Synthesis,” Working Paper No.
752, Economic Growth Center, Yale University, February 1996.
66
Collier, Irwin L., Jr., “On the First Year of German Monetary, Economic and
Social Union,” Journal of Economic Perspectives, 5 (Fall 1991), pp. 179–186.
Commission of the European Communities, “The ERM in 1992,” European
Economy, 54 (1993), pp. 143–157.
———, The Impact of Currency Fluctuations on the Internal Market, Brussels,
COM(95) 503 final, October 31, 1995.
Committee of Governors of the Central Banks of the Member States of the
European Economic Community, Annual Report 1992, Basle, Committee,
April 1993.
Cooper, Richard N., “Economic Interdependence and Coordination of Economic
Policies,” in Ronald W. Jones and Peter B. Kenen, eds., Handbook of International Economics, Vol. 2, Amsterdam, North-Holland, Elsevier, 1985.
Currie, David A., and Paul Levine, Rules, Reputation and Macroeconomic Policy
Coordination, Cambridge, Cambridge University Press, 1993.
De Grauwe, Paul, “Towards European Monetary Union without the EMS,”
Economic Policy, 18 (April 1994), pp. 147–185.
de Kock, Gabriel, and Vittorio U. Grilli, “Fiscal Policies and the Choice of
Exchange Rate Regime,” Economic Journal, 103 (March 1993), pp. 347–358.
Deutsche Bundesbank, Monthly Report of the Deutsche Bundesbank, Frankfurt,
Deutsche Bundesbank, July 1990 and July 1992.
Drazen, Allan, and Paul R. Masson, “Credibility of Policies versus Credibility of
Policymakers,” Quarterly Journal of Economics, 109 (August 1994), pp.
735–754.
Eichengreen, Barry, Andrew K. Rose, and Charles Wyplosz, “Speculative Attacks
on Pegged Exchange Rates: An Empirical Exploration with Special Reference
to the European Monetary System,” National Bureau of Economic Research
Working Paper No. 4898, Cambridge, Mass., National Bureau of Economic
Research, October 1994.
———, “Contagious Currency Crises,” National Bureau of Economic Research
Working Paper No. 5681, Cambridge, Mass., National Bureau of Economic
Research, July 1996.
Eichengreen, Barry, and Charles Wyplosz, “The Unstable EMS,” Brookings
Papers on Economic Activity, 1 (1993), pp. 51–124.
Emminger, Otmar, D-Mark, Dollar, Nährungskrisen, Stuttgart, Deutsche
Verlags-Austalt, 1986.
Flood, Robert P., and Peter M. Garber, “Collapsing Exchange-Rate Regimes:
Some Linear Examples,” in Flood and Garber, Speculative Bubbles, Speculative Attacks, and Policy Switching, Cambridge, Mass., and London, MIT
Press, 1994, pp. 179–191 (previously published 1984).
Frankel, Jeffrey A., “Ambiguous Policy Multipliers in Theory and in Empirical
Models,” in Ralph C. Bryant, Dale W. Henderson, Gerald Holtham, Peter
Hooper, and Steven A. Symansky, eds., Empirical Macroeconomics for
Interdependent Economies, Washington, D.C., Brookings Institution, 1988,
pp. 17–26.
67
Fratianni, Michele, and Jürgen von Hagen, The European Monetary System and
European Monetary Union, Boulder, Colo., Westview Press, 1992.
Froot, Kenneth A., and Kenneth S. Rogoff, “The EMS, the EMU, and the
Transition to a Common Currency,” NBER Macroeconomics Annual (1991),
pp. 269–317.
Garber, Peter M., and Lars E.O. Svensson, “The Operation and Collapse of
Fixed Exchange Rate Regimes,” in Gene M. Grossman and Kenneth S.
Rogoff, eds., Handbook of International Economics, Vol. 3, Amsterdam,
North-Holland, Elsevier, 1995, pp. 1865–1911.
Gerlach, Stefan, and Frank Smets, “Contagious Speculative Attacks,” CEPR
Discussion Paper No. 1055, London, Centre for Economic Policy Research,
November 1994.
Ghosh, Atish R., and Paul R. Masson, Economic Cooperation in an Uncertain
World, Cambridge, Mass., and Oxford, Blackwell, 1994.
Giavazzi, Francesco, and Alberto Giovannini, Limiting Exchange Rate Flexibility: The European Monetary System, Cambridge, Mass., MIT Press, 1989.
Giavazzi, Francesco, and Marco Pagano, “The Advantage of Tying One’s Hands:
EMS Discipline and Central Bank Credibility,” in Torsten Persson and
Guido Tabellini, eds., Monetary and Fiscal Policy, Vol. 1: Credibility,
Cambridge, Mass., and London, MIT Press, 1994, pp. 225–246 (previously
published 1988).
Goldstein, Morris, David Folkerts-Landau, Peter M. Garber, Liliana RojasSuárez, and Michael G. Spencer, International Capital Markets, Part I:
Exchange Rate Management and International Capital Flows, Washington,
D.C., International Monetary Fund, April 1993.
Gros, Daniel, and Niels Thygesen, European Monetary Integration: From the
European Monetary System to European Monetary Union, London, Longman,
and New York, St. Martin’s, 1992.
Hamada, Koichi, “A Strategic Analysis of Monetary Interdependence,” Journal
of Political Economy, 84 (August 1976), pp. 677–700.
Hughes Hallett, Andrew J., and Yue Ma, “Economic Cooperation within Europe:
Lessons from the Monetary Arrangements in the 1990s,” CEPR Discussion
Paper No. 1190, London, Centre for Economic Policy Research, June 1995.
Jeanne, Olivier, “Models of Currency Crises: A Tentative Synthesis—with Special
Reference to the 1992–93 EMS Crisis,” Paris, ENPC-CERAS Working Paper,
Paris, Ecole Nationale des Ponts et Chaussées and Centre d’Enseignement et
de Recherche en Analyse Socio-Economique, June 1995.
Kenen, Peter B., Economic and Monetary Union in Europe: Moving Beyond
Maastricht, Cambridge, Cambridge University Press, 1995.
———, “Analyzing and Managing Exchange-Rate Crises,” Open Economies
Review, 7 (Supplement 1, 1996), pp. 469–492.
Krugman, Paul R., “A Model of Balance-of-Payments Crises,” in Krugman,
Currencies and Crises, Cambridge, Mass., and London, MIT Press, 1992,
pp. 61–76 (previously published 1979).
68
McKibbin, Warwick J., and Jeffrey D. Sachs, Global Linkages: Macroeconomic
Interdependence and Cooperation in the World Economy, Washington, D.C.,
Brookings Institution, 1991.
Marston, Richard C., International Financial Integration: A Study of Interest
Differentials between the Major Industrial Countries, Cambridge, Cambridge
University Press, 1995.
Masson, Paul R., Steven A. Symansky, and Guy Meredith, MULTIMOD Mark II:
A Revised and Extended Model, Occasional Paper No. 71, Washington, D.C.,
International Monetary Fund, July 1990.
Muehring, Kevin, “Currency Chaos: The Inside Story,” Institutional Investor
(October 1992), pp. 11–15.
Obstfeld, Maurice, “Rational and Self-Fulfilling Balance-of-Payments Crises,”
American Economic Review, 76 (March 1986), pp. 72–81.
———, “Destabilizing Effects of Exchange Rate Escape Clauses,” National
Bureau of Economic Research Working Paper No. 3630, Cambridge, Mass.,
National Bureau of Economic Research, January 1991.
———, “The Logic of Currency Crises,” Cahiers Economiques et Monétaires
(Banque de France), 43 (1994), pp. 189–213.
———, “Models of Currency Crises with Self-Fulfilling Features,” European
Economic Review, 40 (April 1995), pp. 1037–1047.
Obstfeld, Maurice, and Kenneth S. Rogoff, “The Mirage of Fixed Exchange
Rates,” Journal of Economic Perspectives, 9 (Fall 1995), pp. 73–96.
Ozkan, F. Gulcin, and Allan Sutherland, “A Model of ERM Crises,” CEPR
Discussion Paper No. 879, London, Centre for Economic Policy Research,
January 1994.
Padoa-Schioppa, Tommaso, “Policy Cooperation and the EMS Experience,” in
Willem H. Buiter and Richard C. Marston, eds., International Economic Policy
Coordination, Cambridge and New York, Cambridge University Press, 1985,
pp. 331–355.
———, The Road to Monetary Union in Europe, Oxford, Clarendon, 1994.
Persson, Torsten, and Guido Tabellini, “Double-Edged Incentives: Institutions
and Policy Coordination,” in Gene M. Grossman and Kenneth S. Rogoff, eds.,
Handbook of International Economics, Vol. 3, Amsterdam, North-Holland,
Elsevier, 1995, pp. 1973–2030.
Rose, Andrew K., and Lars E. O. Svensson, “European Exchange Rate Credibility
Before the Fall,” European Economic Review, 38 (June 1994), pp. 1185–1216.
Salant, Stephen W., and Dale W. Henderson, “Market Anticipations of Government Policies and the Price of Gold,” Journal of Political Economy, 86
(August 1978), pp. 627–648.
Schlesinger, Helmut,, “A Perspective from Germany: The Economy after Unification and a Look at the Progress of European Integration,” The John Foster
Dulles Program Lectures, Woodrow Wilson School, Princeton University,
Princeton, N.J., October 19, 1994.
69
Sinn, Gerlinde, and Hans-Werner Sinn, Jumpstart: The Economic Unification
of Germany, translated by Juli Irving-Lessmann, Cambridge, Mass., and
London, MIT Press, 1992.
Svensson Lars E. O., “Fixed Exchange Rates as a Means to Price Stability:
What Have We Learned?” European Economic Review, 38 (April 1994),
pp. 447–468.
Velasco, Andrés, “Fixed Exchange Rates: Credibility, Flexibility and Multiplicity,”
European Economic Review, 40 (April 1996), pp. 1023–1035.
70
PUBLICATIONS OF THE
INTERNATIONAL FINANCE SECTION
Notice to Contributors
The International Finance Section publishes papers in four series: ESSAYS IN INTERNATIONAL FINANCE, PRINCETON STUDIES IN INTERNATIONAL FINANCE, and SPECIAL
PAPERS IN INTERNATIONAL ECONOMICS contain new work not published elsewhere.
REPRINTS IN INTERNATIONAL FINANCE reproduce journal articles previously published
by Princeton faculty members associated with the Section. The Section welcomes the
submission of manuscripts for publication under the following guidelines:
ESSAYS are meant to disseminate new views about international financial matters and
should be accessible to well-informed nonspecialists as well as to professional economists. Technical terms, tables, and charts should be used sparingly; mathematics should
be avoided.
STUDIES are devoted to new research on international finance, with preference given
to empirical work. They should be comparable in originality and technical proficiency
to papers published in leading economic journals. They should be of medium length,
longer than a journal article but shorter than a book.
SPECIAL PAPERS are surveys of research on particular topics and should be suitable
for use in undergraduate courses. They may be concerned with international trade as
well as international finance. They should also be of medium length.
Manuscripts should be submitted in triplicate, typed single sided and double spaced
throughout on 8½ by 11 white bond paper. Publication can be expedited if manuscripts
are computer keyboarded in WordPerfect or a compatible program. Additional
instructions and a style guide are available from the Section.
How to Obtain Publications
The Section’s publications are distributed free of charge to college, university, and
public libraries and to nongovernmental, nonprofit research institutions. Eligible
institutions may ask to be placed on the Section’s permanent mailing list.
Individuals and institutions not qualifying for free distribution may receive all
publications for the calendar year for a subscription fee of $40.00. Late subscribers will
receive all back issues for the year during which they subscribe. Subscribers should
notify the Section promptly of any change in address, giving the old address as well as
the new.
Publications may be ordered individually, with payment made in advance. E SSAYS
and REPRINTS cost $9.00 each; STUDIES and SPECIAL PAPERS cost $12.50. An
additional $1.50 should be sent for postage and handling within the United States,
Canada, and Mexico; $1.75 should be added for surface delivery outside the region.
All payments must be made in U.S. dollars. Subscription fees and charges for single
issues will be waived for organizations and individuals in countries where foreignexchange regulations prohibit dollar payments.
Please address all correspondence, submissions, and orders to:
International Finance Section
Department of Economics, Fisher Hall
Princeton University
Princeton, New Jersey 08544-1021
71
List of Recent Publications
A complete list of publications may be obtained from the International Finance
Section.
ESSAYS IN INTERNATIONAL FINANCE
171. James M. Boughton, The Monetary Approach to Exchange Rates: What Now
Remains? (October 1988)
172. Jack M. Guttentag and Richard M. Herring, Accounting for Losses On Sovereign
Debt: Implications for New Lending. (May 1989)
173. Benjamin J. Cohen, Developing-Country Debt: A Middle Way. (May 1989)
174. Jeffrey D. Sachs, New Approaches to the Latin American Debt Crisis. (July 1989)
175. C. David Finch, The IMF: The Record and the Prospect. (September 1989)
176. Graham Bird, Loan-Loss Provisions and Third-World Debt. (November 1989)
177. Ronald Findlay, The “Triangular Trade” and the Atlantic Economy of the
Eighteenth Century: A Simple General-Equilibrium Model. (March 1990)
178. Alberto Giovannini, The Transition to European Monetary Union. (November
1990)
179. Michael L. Mussa, Exchange Rates in Theory and in Reality. (December 1990)
180. Warren L. Coats, Jr., Reinhard W. Furstenberg, and Peter Isard, The SDR
System and the Issue of Resource Transfers. (December 1990)
181. George S. Tavlas, On the International Use of Currencies: The Case of the
Deutsche Mark. (March 1991)
182. Tommaso Padoa-Schioppa, ed., with Michael Emerson, Kumiharu Shigehara,
and Richard Portes, Europe After 1992: Three Essays. (May 1991)
183. Michael Bruno, High Inflation and the Nominal Anchors of an Open Economy.
(June 1991)
184. Jacques J. Polak, The Changing Nature of IMF Conditionality. (September 1991)
185. Ethan B. Kapstein, Supervising International Banks: Origins and Implications of
the Basle Accord. (December 1991)
186. Alessandro Giustiniani, Francesco Papadia, and Daniela Porciani, Growth and
Catch-Up in Central and Eastern Europe: Macroeconomic Effects on Western
Countries. (April 1992)
187. Michele Fratianni, Jürgen von Hagen, and Christopher Waller, The Maastricht
Way to EMU. (June 1992)
188. Pierre-Richard Agénor, Parallel Currency Markets in Developing Countries:
Theory, Evidence, and Policy Implications. (November 1992)
189. Beatriz Armendariz de Aghion and John Williamson, The G-7’s Joint-and-Several
Blunder. (April 1993)
190. Paul Krugman, What Do We Need to Know About the International Monetary
System? (July 1993)
191. Peter M. Garber and Michael G. Spencer, The Dissolution of the Austro-Hungarian
Empire: Lessons for Currency Reform. (February 1994)
192. Raymond F. Mikesell, The Bretton Woods Debates: A Memoir. (March 1994)
193. Graham Bird, Economic Assistance to Low-Income Countries: Should the Link
be Resurrected? (July 1994)
72
194. Lorenzo Bini-Smaghi, Tommaso Padoa-Schioppa, and Francesco Papadia, The
Transition to EMU in the Maastricht Treaty. (November 1994)
195. Ariel Buira, Reflections on the International Monetary System. (January 1995)
196. Shinji Takagi, From Recipient to Donor: Japan’s Official Aid Flows, 1945 to 1990
and Beyond. (March 1995)
197. Patrick Conway, Currency Proliferation: The Monetary Legacy of the Soviet Union.
(June 1995)
198. Barry Eichengreen, A More Perfect Union? The Logic of Economic Integration.
(June 1996)
199. Peter B. Kenen, ed., with John Arrowsmith, Paul De Grauwe, Charles A. E.
Goodhart, Daniel Gros, Luigi Spaventa, and Niels Thygesen, Making EMU
Happen—Problems and Proposals: A Symposium. (August 1996)
200. Peter B. Kenen, ed., with Lawrence H. Summers, William R. Cline, Barry
Eichengreen, Richard Portes, Arminio Fraga, and Morris Goldstein, From Halifax
to Lyons: What Has Been Done about Crisis Management? (October 1996)
201. Louis W. Pauly, The League of Nations and the Foreshadowing of the International
Monetary Fund. (December 1996)
202. Harold James, Monetary and Fiscal Unification in Nineteenth-Century Germany:
What Can Kohl Learn from Bismarck? (March 1997)
203. Andrew Crockett, The Theory and Practice of Financial Stability. (April 1997)
204. Benjamin J. Cohen, The Financial Support Fund of the OECD: A Failed Initiative.
(June 1997)
205. Robert N. McCauley, The Euro and the Dollar. (November 1997)
206. Thomas Laubach and Adam S. Posen, Disciplined Discretion: Monetary Targeting
in Germany and Switzerland. (December 1997)
PRINCETON STUDIES IN INTERNATIONAL FINANCE
61. Stephen A. Schuker, American “Reparations” to Germany, 1919-33: Implications
for the Third-World Debt Crisis. (July 1988)
62. Steven B. Kamin, Devaluation, External Balance, and Macroeconomic Performance: A Look at the Numbers. (August 1988)
63. Jacob A. Frenkel and Assaf Razin, Spending, Taxes, and Deficits: InternationalIntertemporal Approach. (December 1988)
64. Jeffrey A. Frankel, Obstacles to International Macroeconomic Policy Coordination.
(December 1988)
65. Peter Hooper and Catherine L. Mann, The Emergence and Persistence of the U.S.
External Imbalance, 1980-87. (October 1989)
66. Helmut Reisen, Public Debt, External Competitiveness, and Fiscal Discipline in
Developing Countries. (November 1989)
67. Victor Argy, Warwick McKibbin, and Eric Siegloff, Exchange-Rate Regimes for
a Small Economy in a Multi-Country World. (December 1989)
68. Mark Gersovitz and Christina H. Paxson, The Economies of Africa and the Prices
of Their Exports. (October 1990)
69. Felipe Larraín and Andrés Velasco, Can Swaps Solve the Debt Crisis? Lessons
from the Chilean Experience. (November 1990)
73
70. Kaushik Basu, The International Debt Problem, Credit Rationing and Loan
Pushing: Theory and Experience. (October 1991)
71. Daniel Gros and Alfred Steinherr, Economic Reform in the Soviet Union: Pas de
Deux between Disintegration and Macroeconomic Destabilization. (November
1991)
72. George M. von Furstenberg and Joseph P. Daniels, Economic Summit Declarations, 1975-1989: Examining the Written Record of International Cooperation. (February 1992)
73. Ishac Diwan and Dani Rodrik, External Debt, Adjustment, and Burden Sharing:
A Unified Framework. (November 1992)
74. Barry Eichengreen, Should the Maastricht Treaty Be Saved? (December 1992)
75. Adam Klug, The German Buybacks, 1932-1939: A Cure for Overhang? (November
1993)
76. Tamim Bayoumi and Barry Eichengreen, One Money or Many? Analyzing the
Prospects for Monetary Unification in Various Parts of the World. (September 1994)
77. Edward E. Leamer, The Heckscher-Ohlin Model in Theory and Practice. (February
1995)
78. Thorvaldur Gylfason, The Macroeconomics of European Agriculture. (May 1995)
79. Angus S. Deaton and Ronald I. Miller, International Commodity Prices, Macroeconomic Performance, and Politics in Sub-Saharan Africa. (December 1995)
80. Chander Kant, Foreign Direct Investment and Capital Flight. (April 1996)
81. Gian Maria Milesi-Ferretti and Assaf Razin, Current-Account Sustainability. (October
1996)
82. Pierre-Richard Agénor, Capital-Market Imperfections and the Macroeconomic
Dynamics of Small Indebted Economies. (June 1997)
83. Michael Bowe and James W. Dean, Has the Market Solved the Sovereign-Debt
Crisis? (August 1997)
84. Willem H. Buiter, Giancarlo Corsetti, and Paolo A. Pesenti, Interpreting the ERM
Crisis: Country-Specific and Systemic Issues (March 1998)
SPECIAL PAPERS IN INTERNATIONAL ECONOMICS
16. Elhanan Helpman, Monopolistic Competition in Trade Theory. (June 1990)
17. Richard Pomfret, International Trade Policy with Imperfect Competition. (August
1992)
18. Hali J. Edison, The Effectiveness of Central-Bank Intervention: A Survey of the
Literature After 1982. (July 1993)
19. Sylvester W.C. Eijffinger and Jakob De Haan, The Political Economy of Central-Bank
Independence. (May 1996)
REPRINTS IN INTERNATIONAL FINANCE
28. Peter B. Kenen, Ways to Reform Exchange-Rate Arrangements; reprinted from
Bretton Woods: Looking to the Future, 1994. (November 1994)
29. Peter B. Kenen, Sorting Out Some EMU Issues; reprinted from Jean Monnet
Chair Paper 38, Robert Schuman Centre, European University Institute, 1996.
(December 1996)
74
The work of the International Finance Section is supported
in part by the income of the Walker Foundation, established
in memory of James Theodore Walker, Class of 1927. The
offices of the Section, in Fisher Hall, were provided by a
generous grant from Merrill Lynch & Company.
ISBN 0-88165-256-3
Recycled Paper ISIS